Peavey Transtube 212 EFX

[gbono]

Found one these on CL with a "for parts" status. With the amp connected to a LB limiter I get a soft orange glow and the channel select button does nothing. FX controls appear to work- LEDs light. No signal from the preamp and when driving the power amp input I get a very weak distorted signal at the output.

I pulled the secondary connectors and get 70VRMS across J6-8 and 35VRMS across J18-22. Reconnect secondary and get +-35VDC instead of +-47V rails. Quick DMM check of filter caps, Q3 and Q6 show no issue. What's loading down the 47V rails?

[Enzo]

probably the bulb.

OK, the bulb did its job, it protected you from shorted circuits, it didn't glow brightly.  So now get rid of the bulb and see if your voltages come back up.

Even if the main power rails are 10v low, the power amp circuits should still work, however, there are op amps in the power amp and in the circuits between the power amp in jack and the amp circuit itself.  And they need good +/-15v to run.  So do you have good 15v rails?

The channel switching IC needs good 15v as well as all the various muting circuits.

Start there, without the 15v rails, nothing else will work.

[gbono]

Thanks for getting back to me.

With LBL removed there is a slight improvement in rail volatge but the +-25V rails are down to 7VDC so the 15 volt rail is even lower. Seems odd that I'm not drawing lots of current - the amp doesn't blow the fuse.

I can only guess that it's Q3/6 though they pass DMM test. Should check actual voltages around EBC. C14/28/15/29?

Looking at the circuit again, I'm wondering if the regulators are damaged? Also is this PA class G - where the 47 and 25V rails are switched depending on output demand?

[Enzo]

I think you are way overanalyzing this.

There are not switched rails, this is a pretty simple amp really.  25v is not a power rail for the power amp stages.  All it does in the power amp is provide current to the zeners, CR8/36, which form the power rails for the power amp op amp U2.  The amp itself just runs on straight +/-47.  Q3/6 are the outputs, Q2/7 are the drivers, and U2 controls them.  Q4/5 are just limiters.  The only thing uncommon here is that this amp is a flying rail type, also called grounded emitter.  Note the center tap of the power transformer is not grounded, it is the speaker lead.  Peavey uses this in most of their amps.

How could Q3/6 pull down the 25v rails?  They run off the 47v rails, and if they were indeed pulling it down, they'd be hot.

Your immediate problem is the 25v supplies, and by extension the 15v supplies that depend on them.  Being loaded down is not the only way voltages can be low.  If the supply is only 7v instead of 25, then what AC voltage is there to rectify?  Look at J18/22, or if it is more convenient, just measure fuse to fuse.  Do you get about 36vAC?  More important, do you get about 18vAC to ground from each?

[gbono]

I pulled the secondary connectors and get 70VRMS across J6-8 and 35VRMS across J18-22. Didn't check to see what the voltages are refereced to ground. Will check this out.

Why drive U2B off the 25V rails with a zener diode/cap regulator when there is a seperate +- 15V rail already used on the other OP Amps?

[Enzo]

Because this is a flying rail amp, and those special zener supplies are referenced to those rails, so the op amp supply for U2 follows them.  The other 15v regulated supplies are for the preamp, and you want them referenced to ground.

You have 35vAC across J18/22 unloaded, but what does it do when connected?  Your concern was the supply was being loaded, so either the voltage stays close to 35v even when loaded, or it indeed drops like a stone from loading.

Just an itch in my suspicions, if the center tap of that 35v does not make it to ground, then the supply won't work.  That is why we want to measure not only the 35v, but also that each end is getting its 18v to ground.

[gbono]

Measured AC voltage to ground for J18 = 19V and J22 = 19V - no limiter connected. The voltage across J18/22 is 38VAC.

When I remove J28 from the reverb board (DSP) the +-25V rails come back to life. Something is frapped on the reverb board - dang.

[Enzo]

OK, that narrows it down.  Plug the board back in and see if any ICs are getting hot.  Risk a fingertip.

[J M Fahey]

I suspect that "Reverb" (EFX) board is an OEM bought, ready made PCB which is not repaired but replaced if necessary.

[Enzo]

No, PV designed and built, schematic in the complete file.  It is 3M file, if I can post that large here, I will.

[gbono]

It's the DSP Package that's getting hot _ I didn't have time to carefully check if any other components (DC converters) are getting hot. Also need to see what other voltages are present on J301....

[gbono]

I found, besides the Cirrus Logic DSP chip, U 300 / LM2594 DC/DC buck converter is also getting hot. I desoldered U 300 and will replace it and see what happens. I'm tempted to also replace C18 though it seems to be okay.

Hopefully it"s the converter since the DSP is not replaceable - Cirrus Logic EOL'ed the CS4826 in 2005.....

[Enzo]

That doesn't mean PV parts department doesn't have a stash of them, did you check with PV?

And while I am not a board swapper, it wouldn't hurt to ask PV if they have thios little DFX board in stock, and the price if so.  it might make sense to replace teh card instead of repair, especially if more than one part has failed.

[gbono]

Here is the update of what Peavey has in stock (they always amaze me):
-----------------------------------------------------------------------------------------------------------------------------------------
Hello George,

99099570----EFX DIGITAL CBD ASSM    Retail $259.99 each (in stock)

70418304----CS4826 DSP/CODEC 2I/6O    Retail $64.81 each (in stock)

The original part number 70300079 has been discontinued. Below is the same with the ROHS number, meaning lead free.
30300079----47UF 35V 20% EL SMT6       Retail $0.49 each (in stock)

Part number 70800037 has been discontinued.

Thanks,
April

-------------------------------------------------------------------------------------------------------------------------------------------

I'm going to replace the regulator/capacitor and test the inductor - too bad it's EOL. Hopefully the issue was the cap drifted and the regulator went unstable.

If that isn't the issue then it would cost +$65 to change out the DSP. I do have access to someone who could rework the 128 pin package reliably and assume you can still host & boot via serial interface (U101 EEPROM - also discontinued). That's more work than I want to do but we'll see.

[J M Fahey]

Problem is: what if you replace the chip and that was not the actual problem?
I hate gambling and this is incomfortably close.

Personally I'd get a guaranteed working bready made multieffects minin board and mount it there somehow.
One example:
http://www.pcm63.com/?232,alesis-al3201-scr-fx-module
for all of U$35  .
With the added bonus that since it's an independent module, if customer does not like it you pull it and use somewhere else.

In fact I thought that's what Peavey used, I have repaired powered mixers where the main construction is conventional through hole parts, which included a small ultra compact SMT board, obviously bought OEM and integrated into the main amp, I have been offered such boards by a supplier for even less than that price (think around 25/30U$ each).

Given the low cost, it's tempting to use them even on a fixed setting, replacing spring reverb.