Great River MP-2 , info, schematics & discussions

[Greg]

Yikes, I'd first "ohm out" the rails to see if a component failed short. If nothing is showing short, it's probably be safe to power it up again and make sure you're voltages are ok. Then you can signal trace down the signal path if it doesn't pass audio.

 

[okgb]

DOUBLE CHECK the original diagram from Dan , but it looks
like the power connections on NRG's diagram may be reversed

i thought i remember the phantom being last , left to right looking
at it from the back of the board , meter drive first ground in the middle

gotta run to a gig will be back to check later today ,

Regards Greg

 

[Dan Kennedy]

The power supply connection on the nrg drawing is reversed!!!

The top pin is the +60v in, next down is +30, then ground, then -30 and the bottom one is the meter feed.

The other connectors are labeled properly.

 

[Bluzzi]

Dan, thank you so very much for steppin' in on that "detail"! Saved us a lot of smoke!

Can you tell me if the meter feed is buffered? I have a bunch of VU meters lying around and wouldn't mind adding them. Thanks.

jim

 

[Dan Kennedy]

The meter feed isn't buffered, but a very simple opamp driver is all you need.

 

[Bluzzi]

Thanks again Dan. JLM makes a VU meter buffer kit that I'll look at.

I just got bench power supply and 'am almost ready to test 2 preamps out.
(I can just see myself now...slowly ramping up voltage as I check for sight and smell of smoke....)

jim

 

[madriaanse]

Okay, here's what I've found so far.

U1 and U2 are putting out 23.5V and -23.5V respectively.

And the preamp actually passes audio and seems to be working as it should gain/amplifier wise, but there is heavy distortion on the negative portion of the waveform. It looks like there's a reverse saw wave superimposed on the signal (1K sine).

Here's a close up of the lower portion of the waveform:

The signal seems to be okay up to 2sk-389 and the left side of D3/D4, but after that it's distorted.

Also, I'm seeing the following on the Q2, Q3 side of 2sk-389

It's the reverse saw wave that is seen on the lower portions of the waveforms (impossible to tell from pic).

And last, I'm unable to trim the output DC offset. Adjusting the trimpot (R46) temporarily adjusts it, but it nearly instantly bounces back to about -20mV

Any insights would be terrific!! Blown 2SK-389?

Thanks,

M.

 

[Dan Kennedy]

No,

I don't think the FET's bad.

But, here's where it's going to get interesting.

Troubleshooting from a distance.

I'm at home, it's late and I don't have all the numbers here, and can't remember them anyway, but I'll ask some questions from work tomorrow.

It's oscillating, obviously, but not in a way I've seen recently. But, and please don't take this wrong, because I probably have seen most ways this can be built wrong, but only once or twice each way, and it's been awhile, but please check all component values.

It's a very high gain, sort of squirrelly amp, as I've made disclaimers before, so proper layout is important.

But, I really do use these things for lots of projects, headphone amps, summing amps, hifi preamps, mic preamps (duh) and occasionally power supply error amps (yep!) and they work just fine.

So my first guess is, what parts are wrong?

 

[Samuel Groner]

When troubleshooting oscillations its very helpful (if not necessary) to note at which gain and with which source impedance it happens, as this might point you in the right direction.

If it just happens at lower gains I'd first check the values of C21, R21 and R12-R15 as these set the main stability margins.

In addition to this it will be clever to check if replacing the gain switch with an apropriate resistor mounted right at the board will solve things.

Samuel

 

[madriaanse]

THanks for the suggestions!!

C21, R21 and R12-R15 all checked out ok. I also hardwired a switch resistor to the PCB and that didn't make a difference. The oscillations happen at all gain levels. Keep in mind that this pre is probably damaged from having the voltages hooked up in reverse. What parts could this have fried?

Many THanks!!

M.

 

[Samuel Groner]

Next you'll need to verify the DC conditions. With input shorted, check the voltage drop across R14, R17, R29, R30 and R32 (that is from Q16 emitter to Q17 emitter). And verify the supply voltage after the regulator. The figures might need some educated guessing if the amp is still oscillating.

No time to calculate detailed numers but for R14 and R17 it should be about 0.65 V, for R29 a tad more and for R30/R32 roughly 100 mV (a bit more is OK, just not 1 mV).

Power supply is about +/- 20 V.

BTW, I do consider it good design practice to have reverse protection diode at the power supply entrance--saved my day at least once...

Samuel

 

[mikep]

I hope this isnt too off-topic, but I have a question about the GR opamp. much respest to dan for putting this out there. Im thinking about building one or two of these, not necesarily for mic amps. Im looking at the schematic and wondering, what is the common-mode voltage range? if you configure the amp as a unity gain noninverting follower, what is the maximum signal level?

thanks,
mike p

 

[mikep]

I hope this isnt too off-topic, but I have a question about the GR opamp. much respest to dan for putting this out there. Im thinking about building one or two of these, not necesarily for mic amps. Im looking at the schematic and wondering, what is the common-mode voltage range? if you configure the amp as a unity gain noninverting follower, what is the maximum signal level?

thanks,
mike p

 

[madriaanse]

THanks Samuel,

I've measured the following:

+V from supply is 30.11V
-V from supply is -30.05V
+Vout from U1 is 23.45v
-Vout from U2 is -23.57
R14/15 Vdrop is .74V
R17/18 Vdrop is .68V
R29 unable to measure voltage here. There's DC voltage that slowly and irregularly pulses here.
R30/31 and R32/33; I didn't measure any voltage drop. no DC voltage present.

The oscillations (reverse saw wave) only occur when the preamp is fed a signal.

Thanks for the help!!

M.

 

[okgb]

the onboard regulation should bring it down to 24 v , so that part
doesn't sound out of place ,
What do the oscillations sound like with a mic ?

Also try checking device by device starting at one end , perhaps the input

good luck & skill

 

[Samuel Groner]

I'm looking at the schematic and wondering, what is the common-mode voltage range?

I think it's not that bad--about 2.5 V below the supplies? Don't expect excellent common mode performance from a JFET input without cascode though. In addition to this, large common mode swings will get the BJT (and the dual FET if you use the Linear Systems replacement which has 10 V lower breakdown voltage ratings) close to or even beyond their voltage rating. There's much to say about the design and I think I'd do quite a few things differently, but it works as shown and the design is proven by many sold units, so what do I say.

If you configure the amp as a unity gain noninverting follower, what is the maximum signal level?

If my above figure is about correct it is more than +25 dBu. But as said before, distortion might become pretty high due to common mode effects, especially with high source impedances.

R29 unable to measure voltage here. There's DC voltage that slowly and irregularly pulses here.
R30/31 and R32/33; I didn't measure any voltage drop. No DC voltage present.

So the output stage has no quiescent current, which makes it slow and causes oscillation at some higher signal level. Now check the DC voltage from Q8 collector to Q9 collector. If that reads seriously below 2.6 V Q12 and/or Q13 are likely damaged (you've allready mentioned this, so this is pretty likely). If that voltage is fine, it's probably Q14 and Q15.

Samuel

 

[Dan Kennedy]

I agree with pretty much everything Samuel's saying.

Here's some measurements from a working board:

R10/R11: 1.03 volts, across R10 makes it oscillate, usually not when measuring R11.

R19/R20: 0.18 volts

R29: 0.81 volts

R30-R33: 0.15 volts

Supplies: 23.4 volts

Keep in mind this thing is almost 20 years old, developed late at night, with beer, not brains.

I would suspect Q's 12 thru 15 took a hit, the outputs are pretty tough.

If anybody wants to re-layout the board, I agree with Samuel's suggestion of reverse protection diodes for DIY'ers, but as a production board, it was pointless, the assembly processes pretty much dictated proper polarities throughout. That and everything wrung out with a variac first time up to protect against assembly error damage.

 

[Samuel Groner]

Thanks for the confirmation, Dan. I wonder if you have any addition specifications you could share on the design besides what's in the manual--e.g. THD at different gains?

Samuel

 

[Dan Kennedy]

Just a quick check of the board I measured:

Input at 23.5dBu 1kHz measured at the Hi-Z jack, from an input transformer driven by my HP8903 gives a THD+Noise result of .0043% at unity gain.

At 24dBu it rises quickly to .89%

The 8903 only does +15dBu max out from it's generator, so I'm using a step-up transformer.

Both cases driving a 600 ohm resistor at the amps output.

I'm pretty much limited by the ambient noise, which is very high, lot's of garbage from lamp ballasts and the like.

The amp is not in a chassis, so the noise shows up everywhere.

At higher gains the distortion stays pretty much the same, just increasing amounts of noise.

I just don't have time today to get more rigorous, but the numbers are pretty much what I remember from years back.

 

[madriaanse]

Eureka! I checked the Q8 - Q9 collector voltage and it was way under, so I replaced Q12/13 and for laughs I also replaced Q14/15 - it seems to be working fine now. I'll go though it later and double check voltages, freq resp. etc. Thanks Samuel and Dan for your help!!!!!!!!!!! :thumb:

M.

PS They Ptown, got any spare 2N4401, 2N4403, MPS8099 and MPS8599's? I had to "borrow" the replacements from my 2nd channel.