Vox VT30 - loud humming

[J M Fahey]

Cool, a new chip, properly mounted, did the trick.
Nothing else to "tinker" with, you got it back to Factory conditions.
Now just plug your guitar into it and play

[jondoe]

Cool, a new chip, properly mounted, did the trick.
Nothing else to "tinker" with, you got it back to Factory conditions.
Now just plug your guitar into it and play

She's not the quietest of girls! I've read some people remark that they are stone dead silent, others have sent them back under warranty for light humming and hiss I wish I had another for comparison

[jondoe]

I'm not sure if I should start a new one but I wanted to update this thread at the very least and continue the saga of this irksome Vox. I've not had a chance to give it a high volume test until recently and I've found that when it's cranked up at high volume all is not well. There is a noticeable saw tooth glitching sound that cuts through the signal at certain frequencies and at higher volume. Cranking the volume up further see's the instances of these clicky/buzzing interruptions increase, from one every second, to many times per second. Hitting some chords on a guitar will cause it to be drowned out buy these noises entirely.

I have checked for DC on the speaker and there is none. I thought perhaps the coil was fried from the previous DC injection, but I can see the glitchy spikes in the test signal at the speaker terminals. Looking on page 6 of the schematic, top left hand corner, EH-3 I think this is the feed from the DSP board (if I unplug it I loose the test signal). I've scoped CN8-1 LINE OUT and found the test signal, which doesn't appear to be glitching, but it can still be heard at the speaker. Given my basic trouble shooting does it sound like the problem exists somewhere on the main power board and not the DSP?

Thanks.

[J M Fahey]

Please post pictures showing us that sawtooth or anythying weird at speaker out, after bit appears alsoscope (and post) power amp input, the idea being to find if itīs generated at power amp itself or it just amplifies faithfully what it gets.

To be more precise, and it being a complex amplifier, we will need 3 screen captures:
1) speaker out
2) TDA2050 input
3) Valvetronix tube emulator thingie
and pray itīs not generated by or inside (3) because itīs a weird circuit and not very well explained by VOX themselves.

[jondoe]

I have a very basic DSO, please excuse the quality



That spike on the waveform are the noises, they don't look sawtooth, but sound a little like it. The number of spikes increase as the volume increases. The same image applies at all tested locations, its visible on 1, 2 and kinda of on 3. However if I remove the line out connector for the DSP board I can see the spikes aren't coming from the board itself, i.e. they aren't visible on the line out when not plugged into the power board.

I removed the DSP board from the equation completely by removing all it's connectors from the power board. I then injected a test signal directly into the line input on the power board at EH-3 and the problem remains, once the input volume gets fairly high, the spikes return.

I think that proves it's not the valvetronix DSP board, but the power board?

[jondoe]

Annoyingly I don't have a spare 4 ohm speaker to test this amp with, but I did try an 8 ohm speaker and cranked it all the way up and there was no spikes/distortion observed. I understand that a 4ohm speakers requires more current than the 8ohm, so the circuit isn't being driven as hard, which might explain the lack of noise. I'm going to try and borrow another 8ohm speaker and run them in parallel, so I can at least rule out the existing speaker.

[J M Fahey]

Ok, this last test explains it.
Sadly you have a fake TDA2050.

Usually a relabelled TDA2030 (which are still in production, at least from second-source suppliers), a fine chipamp used in millions of 15W amplifiers, it simply cannot stand +/-24V rails, specially if driving 4 ohm loads at the same time, period.
It-is-not-a-TDA2050

It can *just* drive an 8n ohm load, and we are not sure for how long.
4 ohms straight trigger the short circuit protection and creates spikes.
In fact, in some chipamps itīs called S.P.I.K.E protection

Even if it does not die instantly, itīs not safe at all.

Answering what you have not yet asked, and assuming you donīt want to junk that fine amplifier, thereīs a couple not so good or somewhat complicated "solutions".

1) replace original speaker with an 8 ohm one.

2) forget TDA2050, no real ones left by now.
Try using an LM1875 which is "almost" a TDA2050 and still in production, it might work and itīs the shortest path.
That said, you might already have one, relabelled TDA2050, which woud explain it standing +/-24V rails but not 4 ohm loads.
But trying a legit one , bought from Mouser or similar, will answer the doubt, one way or another. I would try it.

3) get an LM3886 board and adapt it inside your chassis, I bet you have enough space left. Post a gut picture.
It will be happy with +/-24V and driving 4 ohm speakers any day of the week.

If not space enough inside, you might mount it "outside" , hanging from chassis, you just need to drill 1 or 2 holes to send supply and ground to new board and speaker wires.
Remember LM3886 will need its own heatsink.
**Maybe** you can repurpose the original one, would need to see a couple pictures.

4) Plan B from Outer Space: just to mention it, but involves some surgery: you *can* boost TDA2030 with a couple power transistors who drive the speaker, so chipamp only delivers a few mA , dissipates very little, and so can be used at almost maximum voltage.
This is the single supply version,a split supply one can also be made.

[jondoe]

Hmmm, I feared that might be the case, having pulled apart one of them I can see the die size is very small so that pretty much confirms my fears. Thank you for detailing some options. For reference here's an internal image;

http://www.voxshowroom.com/ct/amp/uk...chassis_12.jpg

It's a bit of a hog to work on with that heat sink in the way

And you're right, I don't want to junk the Vox, the DSP board seems fine Would a TDA2051 be a suitable replacement? There are some available from reputable sellers (unlike the TDA2050), I appreciate it's higher wattage output, but I might be able to get a "real" one at least.

[Jazz P Bass]

The TDA 2051 will work (if it is real).

[J M Fahey]

Go for it.
It will put out exact same power as before, since that is determined by power supply capabilities.

The LM3886 conversion I suggested earlier would also have put out exact same power, even being a "68 W" chipamp.

Your picture tells me VOX was well aware that they were pushing TDA2050 to its limit , thatīs why they bused that HUGE heat sink (for a TDA2050 amp that is).

Trying to extract down to the last Watt datasheet promises is Bad Engineering, because bthe Real World out there is way harsher than white coat Lab technicians know.

"They" use regulated supplies, test amps "in the open", naked over a test bench (hint: perfect ventilation) , not cramped inside a cabinet, shaken by speaker, bumped all the way from home-show-home , dropped on concrete, etc.

dumbassbob just damaged perfectly good speakers feeding low frequency synthesized sounds into a speaker mounted in an open back cabinet and expecting nothing below 83 Hz, go figure.

[jondoe]

I've ordered one from a respected UK audio component supplier so fingers crossed. I see Mouser still list them as available to order but not stocked, so there must be some "real" ones floating around. I watched that video on the fakes and then I tested the short protect and sure enough there is none, it just pops the IC instantly. Even more amusing, all the date codes on the one in the video are the same as the ones I was supplied, so that pretty much confirms it.

I'll report back when I get the 2051 and let you know how I get on.

Thanks for you help.

[Jazz P Bass]

There is a seller in the USA that is offering the TDA2051.
The picture shows the nice wide legs.
No back picture of the Liberty Bell heatsink tab though.


Then there are the bogus chips.
Thin legs, rectangular tab.



I could not capture the image but this seller is a hoot.
Shown in the first image is, I presume, a real IC.
Right down to the Liberty Bell tab.
The next image shows what is actually for sale. A 5 piece lot with a rectangular tab.
https://www.ebay.com/itm/1pcs-TDA203....c100005.m1851

[jondoe]

Well, the TDA2051V's arrived, I installed one and...... the problem persists. The ones I was supplied don't quite look like the fakes ones, but don't quite look like the real ones either.

[jondoe]

Images of the 2051's in question.

[J M Fahey]

Ouch!!

Plan 1: get an LM1875 (which in principle nobody cared to fake) from a reputable seller (or bite the bullet and order a couple from Mouser, to justify postage) , it will handle voltage fine and *may* supply enough current for your speaker.

Plan 2: Please post a relatively close up picture showing the other (frontal) side of the heat sink,where the TDA2050 is mounted.
Maybe there is space enough to add 2 TO220 transistors plus one fake TDA2050 or real LM1875 used as a driver, I can suggest a current booster way simpler than the one I showed before: 2 transistors + 1 resistor

Plan 3 ... which I should try first since itīs so simple and quick: *maybe* the fake chips you got so far can "almost" drive your 4 ohm speaker, or they can, but become unstable, you show a nice sinewave, just marred by the spikes, I suggest you add a 2 ohm 10W resistor in series with speaker, I have a hunch it might solve the problem.
If so, get a terminal strip or worst case nail 2 brass nails to the inside of the cabinet to hold the resistor firm but not touching the wood, so it can have fresh air around.

Donīt worry about the "power loss", in fact that speaker has quite more than 4 ohms over most of the range, and you will only lose some around 400Hz ... which is actually good, and will make amp sound "tubier".