Hello, I currently use a a cheap external enclosure for a Crucial MX250 2.5” SSD Sata drive (powered via the USB cable) I record only stereo tracks to this.
I am about to record 32 multitracks. Is anybody using this type of combination with one of the types of cheap enclosure available from Amazon (UK) - was it necessary to use an enclosure with separate power supply?
I will be testing with all tracks running but I have a wedding shortly & any advice would be helpful.
Many thanks, Ed
Not clear what you’re asking. 32ch shouldn’t need more power, just more bandwidth and throughput.
Once I moved yo the Samsung T7 I’ve never looked back, been rock solid. The few bucks saved can’t be worth the hassle of finding out at the wrong time
If no one else has this setup, just theoretically:
The USB 2.0 port on the SQ delivers 2.5W of power.
Unfortunately, I can’t find any data for a Crucial MX250 SSD, but an MX500 is supposed to consume a maximum of 5W.
If your MX250 should requires similar power, an external power supply would certainly be advisable.
Until now, I assumed that an SSD consumes significantly more power when writing 32 mono tracks than when writing one stereo track.
Is that really incorrect?
Thank you for your replies. I have found an early data sheet for the Crucial MX500 (my error) & that shows the following: Power consumption: 250GB: <3.5W; 500GB: <4.5W; 1000GB/2000GB: <5.0W That’s obviously pretty close to the USB 2 standard. I am going to stress test an inexpensive external drive enclosure with a power supply. I will add that ti this chain. With the very high current prices on memory, options are useful. I will also try some of my many spare spinning disc HD’s.
It’s been my experience that the USB port in the SQ is pretty worthless as a power source unless you are charging a phone. I never, ever, record out of my SQ without using an externally powered drive. After that, I have had no failures with any spinning or SSD I have tried up to 4GB. If you are using a bigger drive, think about how sad you’ll be when the drive crashes with every bit of your recording on it. I run 2GB mostly and use one per show. Get it home and download it. Format it before I use it again. Seems to work well.
D.
As mentioned before, USB 2.0 only provides a maximum of only 2.5W.
The Samsung T7, which has been frequently recommended in this forum, consumes a maximum of only 2.5W, which is why it obviously works very reliably with the SQ.
But all the drives you listed, however, can consume significantly more.
For your project, the drives need to write 32 tracks in real time without any interruption and at least virtually in parallel.
This information could be very helpful to others.
Ah, interesting about the T7. I bought two T5s and they failed miserably on my SQ on large track counts. Do you know if they use more wattage than the T7?
Cute little drives tho.
D.
For 32 tracks at 24 bits 48kHz sample rate, we’re looking at something like 4 to 5 MBps of write rate. It’s not a lot for an SSD.
I know the SQ streams data to the HDD instead of accumulating and writing, so there is no violent write burst. My guess is that it should be okay.
Test it if you’re unsure ! Take a computer, plug it in the USB behind the desk, output a test signal from the computer to the SQ USB driver, patch 32 SQ channels so that they take their input from this USB input, and then you record the direct out of those channels as 32 tracks on the USB HDD.
After your test is done, you can take those samples and use some trickery on your favorite DAW to do a null test on all the 32 tracks. That’s gonna show the slightest glitch.
We can find indications that T5 can consume up to 2.8W.
Whether that’s the reason for the problems in your case, I obviously can’t say.
Up until now, I’ve assumed (also because that’s what I’ve read) that power consumption doesn’t only primarily increase with the data rate, but rather with the continuous parallel writing of many files.
But of course, I could be wrong.
Maybe it can find someone here who could measure it.
One more thought: Perhaps this applies more to USB 2.0 interfaces …
I’m actually using a crucial mx250 and other 250G-SSDs: Samsung, some-No-Name, some Dell-OEMs (Micron?) - all those work either with a cheap USB-to-SATA dongle or a cheap external enclosure off of Amazon. I’ve never worried about power consumption, and, as of yet never had any dropouts.
EDIT: recording 29-ish tracks at 24bit/48KHz - even though data rate probably won’t really matter any way. SSDs will buffer the data and write whenever they feel like it.
Building a current limiting device (even just a fuse) into an USB port increases part count and thus cost. Since the Desks might not be manufactured in tens of millions it could be that A&H has something in there - but it might just as well be that it’s just directly wired to a 5V-Rail of the PSU.
I won’t think about SSD wattage on my SQ6 desk until I notice any problems - but of course if the impact of losing a recording is great enough, you want the risk as small as possible.
OTOH: External power supply for the SSD can fail, too - and if it doesn’t seamlessly fall back to bus power in that case… Go figure.
EDIT: you might want to look into Enterprise/Server/Datacenter grade SSDs with so called “Power Loss Protection” to hedge your bets - as said before, Desktop-SSDs will write Data to permanent storage whenever they feel like, they aren’t as reliable as you’d think or even like them to be. Just as reliable as they statistically need to be - comes down to cost again.
I just did a show, and recorded 32 tracks in two 90 min sets no problem. On a tiny samsung usb3.1 thumb drive. It should just be called a fingertip drive as it’s only 1/2” tall.
Dave’s post is one of the reasons that everyone is so confused about what it takes to feel confident that a storage media will work for large track counts on the SQ. “Why does his thumb-drive work and “mine” fails at the worst moment?” There seems to be no rhyme or reason to what works and what doesn’t. And wondering if what worked yesterday will work today.
I guess that I have made the “go big” solution mine. The drives I use, externally powered, have not failed since I started using them. FWIW.
D.
I did spend some time finding a drive with a very fast write speed. Samsung FIT Plus USB Flash Drive with 300 mbps . available on ebay for $20ish.
Congratulations on finding one!
But unfortunately, the name Samsung, the product name, the seller, the size, the (presumbly) read speed (and even the conveniently omitted write speed) are no guarantee of a working USB drive.
With a bit of luck, it might work if you pay attention to the manufacturing year and the product number (see image).
For example, I also used the list as a guide and bought a “fit” Samsung one (which is 2.36 mm high).
But it didn’t work for multitrack recording as I had hoped.
Right - the issue is that not all Thumb drives are made equal: even when they actually have a USB interface that will do the advertised 5 or 10 (or more?) Gigabit/s that doesn’t say anything about the controller, the cache and the flash chips built into that thing. There are boatloads full of useless junk thumb drives that write maybe 10 MBytes/s if one is lucky and the thing was never used before. Just yesterday I saw a “bigclive” youtube video where he split open a solid looking USB-C SSD drive to find it’s just an adapter and a flakey SD card in there.
So even though SQ6 probably only uses USB2 speeds (because that’s enough), this will work only as long as the attached storage actually delivers on it.
Add heat dissipation and thermal throttling to the equation: the smaller the case, the more plastic and/or rubbery stuff the thumb drive’s case is made from the harder it will be for the components to dissipate heat, and too much heat on the controller etc. can lead to drastic slowdowns…
That (and availability, i.e. I had them lying around) is why my go to medium is regular PC-SSDs with maybe 250GB or something with a SATA-USB-adapter. As described above I don’t think that power will be an issue and one might even go into more detail about power usage profiles of these things to make sure it really isn’t: Measure average use, figure out what kind of usage peaks it produces. It hardly will be “this device will use constant 5W when writing” because SSDs are actually a bit more complicated than “I’m now writing”.
What SSD Drive is this? And what made it a flakey SD Card? Did you stress test it and it failed?
Thats beside the point - when I expect an SSD, I don’t want to get a SD card. See for yourself: https://www.youtube.com/watch?v=sukBSvOTDY0
Yeah well you aren’t going to get anything genuine from the likes of AliExpress/Wish/AliBaba/Temu. Especially if you are ordering something that is disconcertingly cheap.
Of course it contains a dodgy SD card rather than proper SSD components 
I’m surely wrong then. Tbh I didn’t checked anything. I’ve a no name 120Go SSD that was too slow for my laptop, and even maybe faulty, put it in a NVME usb c box and I use it as a multi track recording backup. Laptop is doing the hard work on USB. Maybe I got lucky because of the write rate of this SSD but compared to what the laptop is recording, everything is bit perfect. I wasn’t expecting writing multiple files could increase SSD consumption. Gonna check this and thanks for the insights.