Sunday, 29 November 2020

Exchange 365 stupidity

My current email host is shutting down my account shortly, they did give me plenty of notice, I just haven't had any time to do anything about it. I don't really have any time to do anything about it now, but the issue is forced since I can't carry on working with no email.

Previously I had investigated possible solutions, the ideal would be my own email server with someone like DigitalOcean, but I don't have the time to set that up. So instead I decided to go with Exchange 365 as this is very cheap at only £3.60 a month for unlimited shared mailboxes. I have multiple mailboxes over multiple domains, and I am the only 'user', so this works out far cheaper than other providers that charge per mailbox.

But of course, with it being an MS product, getting it set up wasn't simple. I do have to give them credit for their Email migration tool though, this was simple to set up and worked well (as far as I can tell). The limit of 35MB per message does seem rather stingy, and wouldn't suit some people, but max message size with my current email provider is far stingier than 35MB, so it was no problem for me.

The first problem I had was setting up my account in Thunderbird (the email client I use). Info I had read said I needed to use the ExQuilla (paid) add-on. Thankfully you can install this without paying, as I couldn't find any way to get this working with my shared mailboxes.

After reading more info, it seemed to be the consensus was that you should setup the accounts in Thunderbird using IMAP rather than using Exchange. So I got my first mailbox set up using the guidelines here: Office 365 Exchange Online - How to access a Shared Mailbox using Thunderbird or other IMAP client, but I couldn't send and was getting something like unknown error sending of the message failed due to unexpected error 80004005 when sending a test email.

Searching the web about this brought up various things. One was about setting up an 'app password', but Microsoft's own articles about how to do this give instructions that are not possible to follow or contain links that go to dead ends. I would guess either MS have moved the location of where an app password can be set up and not updated their documentation, or it is something for personal MS accounts only, or its been removed.

I managed to turn off 2FA for my account via the Azure AD web admin: https://portal.azure.com/ > Azure AD > Users > Multi Factor Authentication (link along the top) > check the user > disable (under quick steps). BTW if you click the 'service settings' tab, here is where you can check that 'app passwords' is enabled, which it was for me already, but as I mentioned, there doesn't appear to be anywhere you can create an app password.

In any event, this had no effect. I tried using the domain.mail.protection.outlook.com address as the SMTP server, but still got nowhere.

So I tried sending an email via powershell, first with the domain.mail.protection.outlook.com address (and port 25), then with the smtp.office365.com server and my main 365 account:

$Cred = Get-Credential
Send-MailMessage -From 'me@myaccount.onmicrosoft.com' -To 'adifferentemail@somewebsite.com' -Subject 'Test mail' -Credential $Cred -Port 587 -SmtpServer 'smtp.office365.com' -UseSsl $TRUE

When the Get-Credential prompts for your user details, you just enter the user and password for your main O365 a/c. This was actually very helpful as it gave me a much more useful error message:

Send-MailMessage : Mailbox unavailable. The server response was: 5.7.606 Access denied, banned sending IP
[my.ip]. To request removal from this list please visit https://sender.office.com/ and follow the directions.
For more information please go to  http://go.microsoft.com/fwlink/?LinkID=526655 AS(1430)

After doing this, you are told it will take 30 minutes to get your IP removed from MS' blocklist. However, after 30 minutes I was still getting the same message. So I tried again, but this time from my shared mailbox address, requested the unblock again (using my shared mailbox email address), and after another 30 minutes I was a bit further. I was now getting:

Send-MailMessage : The SMTP server requires a secure connection or the client was not authenticated. The server
response was: 5.7.57 SMTP; Client was not authenticated to send anonymous mail during MAIL FROM

This didn't make any sense, as I wasn't sending anonymous email, I was providing the credentials. After much more searching I found a solution that fixed it - the advice was to go to Office 365 > Users > Active Users > Click on the user name > Mail tab > Manage Email apps > Ensure Authenticated SMTP is checked. Now in my case it was ticked, but after saving the settings (despite making no changes) I could now magically send the email. I have since turned back on 'Enable Security defaults' in Azure AD (one piece of advice was to turn this off) and re-enabled 2FA, and it is still working.

So it seems the final setup is:

  • Create shared mailbox, add an alias to it and ensure your main account has full access to it (which it should be default).
  • Go to your user in Office 365 admin, and save the 'Manage email apps' setting (ensuring that Authenticated SMTP is checked).
  • Settings for IMAP Configuration:
    EMAIL ADDRESS:
    shared@anotherdomain.com (shared mailbox)
    IMAP SERVER:
    outlook.office365.com 993 SSL
    USERNAME:
    user@maindomain.com\shared-mailbox (user\shared mailbox alias)
    PASSWORD:
    main user's password
  • Settings for SMTP:
    SMTP SERVER:
    smtp.office365.com 587 TLS
    USERNAME:
    user@maindomain.com (main username)
    PASSWORD:
    main user's password

Additionally, it seems that you actually get users created for each shared mailbox, and you can use these accounts to directly login and send email. But I'm a bit confused on this if the account is just having a trial period, and after that only the main user account will be able to access the shared account. It would be weird if for £3 a month you actually got unlimited users and mailboxes (up to 50GB total storage), just that all mailboxes must be set as shared.

Oh yes, another annoyance was when setting up the account on my phone (Gmail app), I added the account as an Exchange a/c (not IMAP) and used the email of the shared mailbox but credentials of the main user. But the actual mailbox it added was that of the main user (it showed the email address of the shared mailbox though). Adding it again using the credentials of the (unlicensed) user for the shared mailbox did work.

When signing up for Exchange 365, I also found it rather disingenuous that they described it as being reliable - they are almost constantly having problems with Exchange 365 services. Thankfully most of the time the problems aren't with actual email delivery, but sometimes that does go down as well (usually they have it back up after a few hours).

So far I've only migrated one account, so I've now got the rest of my accounts to migrate. Hopefully I've now got things sorted so the rest will be smoother.

I also almost forgot to mention that many of the answers.microsoft.com forum threads that came up in the search results would just load with We are sorry, the thread is not available at this time. Basically everything MS is broken.

Tuesday, 1 September 2020

Vagrant Read-only file system fix

When I booted up my main VM today, it would boot part way and then I'd get:

==> default: Configuring and enabling network interfaces...
C:/HashiCorp/Vagrant/embedded/gems/2.2.9/gems/net-scp-1.2.1/lib/net/scp.rb:398:in `await_response_state': scp: /tmp/vagrant-network-entry-1598986668: Read-only file system (RuntimeError)

Followed by a load of errors. When accessing the machine via Virtualbox, I would also get quite a few read only file system errors after logging in.

Eventually I found the solution was to run a file system check (accessing the machine via Virtualbox):

sudo fsck -C /dev/sda1

(Where /dev/sda1 is the filsystem partition). This found numerous errors, which I just had to put 'y' in to let it fix them. It ran pretty quickly, probably done in a minute. After that I rebooted the VM, logged in and didn't get any errors. Shut it down, then booted it normally via vagrant up, and all working so far.

A good explanation of why the filesystem may become read-only is here: Ubuntu VM “read only file system” fix?.

Saturday, 25 July 2020

Avantree Audikast Plus bluetooth audio transmitter

I recently purchased a Avantree Audikast Plus Wireless Audio Transmitter for around 50 GBP from Ali Express. This device offers AptX Low Latency support and can accept a 3.5mm stereo input (as well as optical and USB audio). So you can use it with practically any device that outputs audio to convert it into bluetooth audio.

My main thought was that with AptX LL support, it might be useable plugged into the headphone socket in a mixer to allow the use of wireless headphones. Now, I'm not a DJ, so I can't really report on whether it is 'usable' for this. Latency seemed low to me, but I couldn't really tell any difference between my headphones (Plantronics Backbeat Pro 2) connected to the Audikast with AptX LL vs the headphones connected to the laptop at the mercy of whatever bluetooth codec windows uses, which I think is SBC.

I thought the Audikast might also work over a longer range than my laptop bluetooth, but actually there was no difference. If there was any difference, it was actually in favour of the laptop's built-in bluetooth. It is considerably better than the 'Bluetooth 5.0' usb dongle I have in my PC though.

Using USB audio, I had a similar problem to the Cyboris X3-PRO-N bluetooth speaker in that the first 250ms or so of sound is cut off. Once an audio signal is going, then it's fine, it's just going from silence to sound you get the problem. When no audio is being output, the input selection light on the Audikast flashes on 'Opt' (optical). Once sound starts playing over USB it will then switch to USB, and stay there until about 15s of silence have passed. At which point it goes back to flashing opt.

Unfortunately there is no way to manually choose the input source. I'm sure they could have included this, e.g. press + and - buttons for 5s, use + and - to select the input you want, with no input selected being auto, then after 5s it sticks on that setting.

Unlike the Cyboris speaker though, when you open Ableton, it switches to USB audio automatically. So no problem with sound being cut in ableton.

With avantree mic selected as audio input in ableton, audio quality for sound was terrible. After turning that off, sound quality was good and no noticable latency.

The Audikast Plus supports 2 sets of headphones connected at once, which could be useful for 2 people wanting to watch TV with headphones. But I don't think I'd ever need this.

Sunday, 14 June 2020

Cyboris X3-PRO-N Wireless Speaker info

I've been after a decent bluetooth speaker for quite a while. Recently on HotUKDeals there was a 'deal' for an Anker Soundcore improved bluetooth speaker, which get good reviews, however one of the commenters suggested that a Cyboris 40W bluetooth waterproof speaker was a better deal. So I decided to purchase one from Ali Express in the hope it really was as good as the commentor made out.

I've now had it for a (short) while, and it is a decent speaker. However, it does have a few quirks, and the manual also doesn't seem to be available online. So I thought I'd document the quirks and post a PDF of the manual I've had to make. On Ali Express it's sold as Upgraded Cyboris 40W Subwoofer Waterproof Portable Bluetooth Speaker NFC TWS Bass Speakers DSP Support MIC TF/Aux Non Tronsmart, however the actual manual calls it X3-PRO-N. Manual download: X3-PRO-N bluetooth waterproof speaker manual.

On bluetooth with audio device set to "Headphones (X3-PRO-N-Stereo)" sound quality is good and sound clips play okay. Latency is bad.

On bluetooth with audio device set to "Headset (X3-PRO-N Hands-Free AG Audio)" sound quality is poor and sound clips play okay. Latency seems improved.

On USB cable (audio device set to "USB Audio") sound quality is good, but the first 250ms or so of a clip is cut off unless the speaker is already playing audio. Latency is good. Same with both DX and Wave.

Playing via audio cable no issues.

Playing from micro SD card plays mp3 and wma files, but not .m4a, .ape, .mpc, .m3u (playlist). Max card size 64GB. Likely has to be formatted ex-FAT or FAT32. Will play music within subfolders nested several deep. Play order is roughly alphabetical, but not quite. In testing it played files in the root, then "Daft Punk" folder, then "Air" folder, "Playground", "Röyksopp", "Wagon Christ", and finally "Vitalic". Files in folders are played first before playing files in subfolders.

When turning on it always goes into bluetooth mode. It did remember the pause position when paused while playing from micro SD card, turned off, then on again.

With the speaker connected to a computer over USB, Generating white noise at 0.001 with Audacity, and playing this gives enough of an audio signal to the speaker that it won't cut off the start of clips when playing them. Using with ASIO4All, it gives exclusive access to the USB Audio device (presumably because it has no ASIO drivers), even when this option is unchecked in the audio device settings. This means you can't play a looped very quiet white noise in one application while using Ableton Live with ASIO 4 All to output to the speakers.

It can't play from a usb stick plugged into the USB port.

Wednesday, 27 May 2020

Heliopan stepping ring

I recently purchased a Heliopan step-up ring to enable mounting filters with a 67mm thread on a lens with a 40.5mm filter thread. It was quite expensive compared to the cheap Chinese stepping rings I normally buy (around 5-10x the cost). When it arrived I was surprised how heavy it was. In looking into this, apparently Heliopan stepping rings are made from Brass (apart from some very large ones), which is much heavier than the aluminium used by the majority of stepping ring manufacturers.

Brass filter threads are meant to be less susceptible to binding than aluminium threads. I have had filters / stepping rings get stuck several times in the past. But for me, it's not worth the extra cost of a brass stepping ring to avoid this relatively rare event, I'm not sure I'd even say it was a worthwhile trade-off for the increased weight. In the case of this particular stepping ring I purchased though, the Heliopan was the only one I could find.

Tuesday, 3 March 2020

Managing volume

Listening to some music, I realised I have 5 volume controls, all completely independent of one another:

  1. The volume control in the audio player
  2. The OS (Windows) main volume control
  3. The USB / Direct mixing knob on the audio interface
  4. The volume knob on the audio interface
  5. The volume knob on the amp

It's a bit too much control really, as if the sound is too quiet there are 5 places you need to check the volume setting.

Saturday, 29 February 2020

Fixing sliders & USB port on M-Audio Axiom Pro 49 keyboard

Last year I purchased an M-Audio Axiom Pro 49 keyboard from eBay. Although I had several 'keyboards' already, I didn't have a 'normal' midi keyboard, which I hoped this would work as.

The keyboard was described (on eBay, where I purchased it from) as being in good condition, but with two broken sliders. However, when I received it, the case was actually quite different (as unfortunately often happens with things bought from eBay). The USB port was very loose and to get the keyboard to work with my PC I had to hold the USB cable in at a certain angle, and stick something heavy on the cable to keep it in place. Some of the keys also make quite a 'croaky' noise when pressed, but they do seem to work okay.

Opening up the keyboard wasn't too difficult, but there are a lot of screws to undo. You don't need to undo the large black screws in the middle of the keyboard's underside, just all the other screws. With the screws undone you can remove the top of the keyboard.

After that you need to remove a few screws holding down a long metal bar that goes across the top of the keyboard. Then you can unscrew the board with the sliders, undo the ribbon cable that attaches it to the main board, and remove the slider board.

I unsoldered one of the broken sliders and found it was an Alpha branded B10K 45mm (30mm travel) slider pot. However, I couldn't find the exact same model anywhere. The nearest I could find is the RA3043F-20-10EB1-B14, which looks to have a plastic rather than metal lever. This is also available on Amazon: Taiwan Alpha Electronics RA3043F-2010EB1-B10K 10k8486; 1/5W PC Mount Linear Taper Slide (Pack of 3).

Although there are plenty of cheap 45mm length (30mm travel) B10K sliders on Ali Express, they all seemed to have the contact pins on the outside of the slider, and also be missing the two inner PCB mounting pins. The Alpha sliders used on the Axiom board have their contact pins a bit further in from the ends of the slider. The only suitable ones I could find on Ali Express were ALPS branded ones, e.g. RS30 Series ALPS Switch 45mm 4.5cm Mono B10K Slide Potentiometer Speaker Mixer Faders Shank Length 10MM, you can find more at the top of this page of search results, and the end of page of the same search results (at the time of writing): 45mm B10K search results page 3.

From eBay I managed to find a set of 4 unbranded ones that look correct: B10K Slide Potentiometer 30mm travel single unit 45mm Length short shaft x4pcs. These were $15 + $2 P&P, and what I purchased. The ALPS ones from Ali Express were $4.5 + $3 P&P ea. The Alpha ones from Amazon were $10.81 + $4.17 P&P. The Alpha ones from Mouser UK were only about £1.5 each, but they didn't have any in stock, and I suspect P&P would probably add quite a bit to the cost as well. The ideal purchase would have been this ebay listing for Alpha 45mm B10K 10K Linear Taper Slide Potentiometer, as that looks to be exactly the same model as is used on the Axiom slider board. However, they only had 2 in stock (which is the number I needed, but I preferred to buy from a source where I could order more in the future should the need arise). They are also much more expensive at $7 + $3 P&P each.

Getting the main board out to look at the USB port was a right pain, it's extremely annoying that the ebay seller didn't mention the broken USB. I only paid £15 for the keyboard though, so I didn't think it was worth sending back for a refund. To get the main board out you have to remove all the various ribbon cables that go into it, and they were very tight and difficult to remove. On one a wire end popped out of the connector as well (no, I wasn't tugging on the wires to pull out the connector blocks).

With all the connectors removed (and the board unscrewed) I could check the USB connection and see that the part with the contacts actually moved quite a fair way back when you pushed it. The female USB B connector block has 4 pins centred on the bottom, so I bought a pack of 10 of these of eBay for £2. I also bought a pack of 10 new slider knobs, since I will need to replace all the slider knobs to keep them matched. So the cost of the sliders, slider caps, and USB connector is actually more than the actual keyboard cost (though less when you include the postage cost on the keyboard).

When trying to find the replacement sliders needed for the keyboard, I found a video about repairing the Axiom 61 keyboard. That said that the individual sliders weren't available, and you have to purchase a complete slider replacement board from M-Audio. They said the cost was only $20, which works out much cheaper than buying the separate sliders anyway. However, it seems you can no longer purchase this board, and the places online that have it listed (but out of stock) charge around $60 + P&P and that price would also attract 20% VAT and £10 post office fee to import into the UK if it was in stock for purchase.

After receiving the sliders and USB connectors I replaced these okay. But the keyboard required the separate power adapter to be plugged in for it to work, when it is meant to be able to take power over USB. Investigating this I found that a component had been ripped off the board behind the USB connector. And I had no idea what type of component this was.

M-Audio's support was completely useless, telling me that I needed to get a replacement main board, which are no longer available. Thankfully I managed to get some help over at electronics stack exchange, where the missing component was identified as probably a ferrite bead. So bought several different values of those (not knowing which value was needed), making this keyboard even more expensive.

When looking to purchase the ferrite beads, I was first looking at Ali Express, but they were quite expensive. I then looked on Mouser, where you could purchase them individually very cheap. But postage was very expensive. Finally I remembered RS Components, a UK based company. Some of the beads on there were only available in very large quantities, but most were available as packs of 10 - 25. While I was looking I also noticed they had the correct slider for the keyboard, and at a decent price (I think this one: Alps Alpine 1 Gang Slide Potentiometer, ±20%, 0.2W Power Rating, Linear, Through Hole ).

So after disassembling and reassembling the keyboard many times, a large amount of time spent looking for the correct parts to fit (and quite a bit of time fitting them) I now have a working keyboard. The only thing I'm still waiting for is the caps for the sliders, which I think must be lost in the post. So I might have to buy those again.

Moving my PC to another case

About a couple of weeks ago one of the drives in my RAID array failed. I managed to find a similar drive to replace it on eBay, though not the exact same model of drive. From my research, so long as the drive is the exact same size (or larger) and same speed (RPM) then it should work fine.

I posted on this blog previously (Why did I do this?! (RAID)) about how I'd fitted the drives into my case (Corsair Air 540). When the replacement drive arrived last week I noticed that the drive temperatures being reported in the RAID console were around 50°C on the drives. And when I took the 'drive cage' out of the case they were indeed pretty hot.


My 'drive cage' (top right) in the rear of the Corsair Air 540 case

Although they had a fan in front of them and two behind them (push - pull configuration) this obviously wasn't enough. I suspect it's that there's not really anywhere in this half of the case for the fans to pull air in from. The front fan is on the rear of a backplane, so most of it is covered by the backplane circuit board, and then the backplane just has a few small holes in the front drive bay doors for air to get pulled through. (I don't normally have any drives in the backplane, so it's not also dealing with having to pull the air past hot drives before it even gets to the drives in my 'drive cage').

All 3 fans cooling the drives were also quite small, around 80 - 90mm. So when I swapped out the dead drive I temporarily ran the drives outside of the case with a couple of spare 140mm fans to keep them nice and cool. This actually worked pretty well.

But it's not really a permanent solution. I managed to procure an old 5U Server case from work, this can be used in rack or tower configuration and is made by PCI Case. In terms of total volume (space) it's probably around the same size as the Air 540. But it's longer and shorter. With a bit of modding, this allows mounting drives at the front of the case with 140mm fans in a push pull configuration in front of and behind them to keep them cool.

Everything else fits in nicely, though I had to swap my Quadro and RAID card slots. While I can't remember off the top of my head, I'm pretty sure the lowest PCI-e slot is faster than the middle slot, so it's a bit annoying to have to swap the cards over. But the bottom slot is right next to the bottom of the case, so the graphics card's fan wouldn't be able to pull in air very well if positioned in the bottom slot.

There's no provision for cable management on the case, so it's a case of zip-tying everything. It took me ages to get the cables looking reasonable, it was a right mess before!

The front 4 fans and the fan on the rear of the backplane I put into a fan splitter. After now having used the machine for a bit, I want to swap a couple of the fans. The pull fan behind the hard drives (on the right) is a Noctua Industrial fan, which I thought would be good placed there because it's so powerful. However, it can be quite noisy, and since it is on the fan hub, it's on the same PWM signal as the other front fans. So I intend to swap that with the rear case fan.

The push fan in front of the hard drives (not visible in the pic above) is a Corsair 3 pin fan. So that's also a bit noisy. I had to use this fan here as the Noctua fans with their square design won't fit in the same position as they block a peg on the front panel. I intend on getting a circular Noctua fan (speed and airflow specs look to be the same as the other fans I'm using, which are NF-AF14 PWM models). Just need something else to buy from Amazon so I can meet the £20 free postage limit. I guess I could get 2 of the circular fans to give a consistent look at the front of the case but I don't care too much how it looks.

The case only has one 3.5″ bay, the rest is all 5.25″ bays. I 'borrowed' a load of adapters from work, which I think are actually meant for floppy drives, but they worked for mounting the hard drives. It gives a lot more space between each drive than they had previously in the DIY 'drive cage'. For the RAID drives I mounted one in the 3.5″ slot, then 3 in the 3 5.25″ bays above that.

In the middle of the case is my backplane, which takes up 2 5.25″ bays. Then there's another single mechanical drive (connected to the motherboard SATA) below that. In the bay below that are 2 2.5″ SATA SSDs, each screwed in just at one edge of the bay. Then the area below that is empty.

You can also see in the above pic one of the smaller fans (that was previously cooling the hard drives in my old case) resting on the rear left corner of the case. This is screwed into side panel to pull in air towards the graphics cards. There aren't actually any mounting holes for the fan, but like in the previous case you can just screw it through the ventilation holes in the panel.

The rear fan I had to drill new mounting holes for as the mounting holes are for a smaller 120mm fan. The fan the case comes with is molex powered, no PWM. The ventilation holes just about cover the area of the 140mm fan though, and since 140mm fans are what I have, it made sense to make it so I could mount a 140mm fan there.

Drilling the holes took absolutely ages. I had also put my motherboard in before I remembered about drilling the holes, so I then had to take it out again (don't really want metal shavings on the motherboard). I also had to take out my motherboard again a 2nd time later - when I had the PC all ready to boot, I was plugging everything in and noticed one of the annoying bendy pieces of metal on the IO shield above a USB port had got bent down and gone inside the USB port. Very annoying and time consuming.

The only other thing to talk about from the pic above is the mounting method for the push pull fans in front of and behind the 5.25″ bays. Currently the front fans aren't secured, but once I've got the fans sorted properly I will cable tie them in place (there's no other way of attaching them).

The fans behind the drives are cable tied onto what I believe are some more 3.5 - 5.25″ bay adapters. They are only attached to the 'top' adapter. The top adapters are then attached to the 'top' of the case with twist ties. This allows undoing them relatively easily, so you can take the adapter off (with fan attached) for access to the drives hidden behind it. I had to drill an extra hole in each of these adapters to allow cable tying the fans to them at the wanted position. Again, this took ages as the metal is relatively thick and I don't have any specialist metal drilling equipment.

The same adapters are used 'below' (in rack mount configuration) each fan. These adapters are cable tied to the case, but the fan is not cable tied to the adapters. They just give a point for the fan to rest against. The adapters provide just enough extra distance for the fan to work without hitting the hard drive power and SATA cables. The bottom adapters are also used for routing cables through.

The above is a general top-down view of the interior of the case. I decided to go for 2 PCI-e 8 pin power cables to the graphics card, not sure if it'll make any difference from having a single cable plugged into both 8 pin sockets as I had before, but it can't really be worse.

This pic shows the fans behind the 5.25″ bays and the rear of the backplane in the center. Trying to manage all those SATA cables was quite tricky. The USB from the backplane doesn't actually go anywhere at the moment - it's not long enough to reach the USB headers on the motherboard, so I need to get an extension for it.

The BBU for the RAID card I just stuck to the case using some double-sided tape.

Looking at the front of the case you might spot a problem - no I/O. I may try and get a front IO panel that fits in the 3.5″ bay slot, but I'm 95% sure there isn't enough space for that due to the hard drive that is mounted in that slot behind the cover. The case does have 2 USB2 ports on the side (top in rack mount). But no headphone jack. So I've had to get out my M-Audio M-Track 2x2M audio interface for audio (this plugs into the PC via USB). I also have 2 USB3 hubs on my desk with the wires connected back to the rear of the PC.

The case does come with 5.25″ bay covers that are slotted / slatted with holes. They would fit in front of the fans okay, but they are more plastic than they are hole. Similarly, there is a front door for the PC with strips cut across it, but again it is more plastic than hole. So I prefer to leave both off for maximum airflow. When my sister's kids come over I can put the door on and lock it so they don't stick stuff into the fans. They only watch Peppa Pig on youtube anyway, so it's not like it should be needing lots of air when they're using it.

Another annoying thing that happened when I was putting the PC together in the 'new' case is that one of the SATA connectors (plastic bit) on a hard drive broke off. I've never had that before. I superglued the connector back on, though the pins don't sit flush against it. I managed to get the cable plugged back on though.

Ideally I'd like to get the Fractal Design Define 7 XL case that was recently released and a lot of youtubers have been giving positive reviews. It's by no means the perfect case for me, but it's probably the nearest to what I'm looking for that's available. Unfortunately it's pretty expensive, and as it's only just released you can't get a used one cheaper. Hopefully at some point I'll be able to do a completely new PC build using that case, but obviously a new PC will be pretty expensive. So I'll just manage with what I've got for the time being.

Saturday, 4 January 2020

AMX, variable AC-DC adapter, and USB hubs

I have read in several places that the audio output of the Akai AMX isn't very loud, and that this can be improved if plugging it into a powered USB hub (the AMX is bus powered). I have two USB 3 hubs, both of which have a socket to plug in a power supply, but I didn't have a 5V DC power adapter that would fit either.

So I purchased a 3-24V Regulated AC-DC adapter (model JDT-2402000 / JDT-001), and a selection of plug adapters. Checking with a multimeter and no load, it seems it goes from about 3.6 - 24V. It displays the voltage on a 3 digit 7 segment display, though I found the voltage measured with the multimeter was always slightly higher than that indicated on the unit.

So I went to test the AMX with my computer to compare volume plugged into the PC, the USB hub unpowered, and the USB hub powered. However, I found that whenever I connected the AMX to a USB hub or the front USB on the PC, Serato (the software I'm using with the AMX) would just say AMX hardware connecting. This can take up to one minute. and never actually connect to the AMX. I tried 3 different USB cables.

Windows detected the AMX fine, but won't output sound via the AMX. Ableton Live will so long as you choose MMX/DX as the driver type and select AMX Wave (not DX) as the output device.

But when I plugged the power adapter into the USB hub, it showed up 4.99 on the adapter's LED, despite the adapter not even being powered on at this point. I adjusted the voltage dial up on the adapter, until I got to 5.02V, anything below this it still showed 4.99V.

With the power in or out there was no difference, the sound output from the AMX was exactly the same volume. This was with a Sumvision USB 3.0 Hub. With my other USB hub, an ipTIME UH305, the voltage with the power adapter off showed as 3.90. But with the power adapter off or at 5V, the volume seemed to be exactly the same.

So I'm not sure where this thing comes from about plugging the AMX into a powered USB hub for higher volume output. With my setup the AMX doesn't even work at all with Serato when plugged into a USB 3 hub, whether the hub is powered or not. And when using the AMX with Ableton, having the USB hub powered or not made no difference.

Maybe it depends on your hardware? Or what else you have plugged into the AMX (I didn't have anything plugged into the audio inputs).

Anyway, although the JDT-2402000 / JDT-001 power adapter isn't useful for what I bought it for, it still seems like a handy piece of kit to have around, especially with all the extra plug adapters I bought to accompany it. The only thing I don't have is an adapter to change it from positive tip to negative tip. However, I should be able to DIY one of these by cutting up a 5-way guitar pedal power adapter cable - just cut each end off, then reconnect them with the wires swapped over.

It would also be better if it did more than 2A (so you could use it power a laptop), but you can't have everything.


The power adapter, variety of tips, and AMX