| jamesm88 |
hi everyone
im new here, and arent really familiar with MOSFET amps, as im a tube guy, but bear with me :)
ive got a perreaux 3000b amp that needed fixing, as the right channel distorted. THAT was fixed easily enough after a bit of dusting and fuse swaps. the left channel however after the fuse swaps now tends to catch fire when a load is connected.
the first fuse holder, the one straight from the power point was a bit bent or stretched or something when i put the a fuse in, so there wasnt an excellent connection, and the 180R resistor inline with the diode coming from the big filter caps and the 270R resistor coming from the emitter of 2 BJT transistors earlier in the circuit went up in smoke. i replaced both, and now the 270R one keeps burning, but the 180R one is fine. id post a schematic, but perreaux dont seem to like handing out copies, but if im desperate, i can email it to you.
any ideas? im at my wits end with this haha. also, there are 2 trimpots in the circuit. how should these be set?
thanks ALOT
james |
|
|
| anatech |
Hi James,
We will need to see the diagram. I have repaired a couple Perreaux amps over the years, but those were ages ago. Parts from Oz are expensive here due to shipping costs. ;)
You could email me a copy at bhomester at gmail dot com . There is most definitely another problem in there.
-Chris |
|
|
| jamesm88 |
good news!! the fires have stopped :D :hot:
NOW i believe i have got something misbiased. i think that because i get a BRUTAL buzzy distortion, and i only get that when i bash the strings hard (im using a bass to test it, as thats what it gets used for). ive come across this in making guitar effects, and i know that it can be fixed with an easy tweak. problem is, im not sure what to tweak.
i managed to stop the resistors burning by replacing a stuffed transistor. i replaced both the transistors in the pair, just in case. the originals were labeled PLSN - not much help, so i took a punt and figured they were just common small signal NPNs, and replaced them with a pair of BC458s. i happen to know that they are pretty high gain transistors, and i also had to make a 'close enough' resistor because i burned all my 270R ones. i currently have a pair of 560R in parallel. they come from the emitters. is this enough difference to affect the bias? |
|
|
| srh |
Hi James,
IIRC the PLSN is a house coded MPSA42 and the PLSP is a MPSA92.
Cheers,
Steve. |
|
|
| anatech |
Hi James,
As Steve posted, there is a large difference in Vce. Replace them right now before more damage occurs. If you have to get the transistors, get the proper resistors (your current value should be close enough)
RV1 and RV3 set the bias current, RV2 and RV4 set the DC offset. Match the current draw of the blown stage to the good stage. If you call Perreaux they should be able to tell you over the phone.
Hint: Turn the amp off and discharge the filter caps through a resistor after letting it sit a bit. Replace one fuse on each side with a 1R0 or 0R1 resistor, I like to use 2W or better. Measure the voltage drops.
-Chris |
|
|
| jamesm88 |
alright the fires are back. i swapped the transistors out as soon as i read your post after a brief moment of panic. i put the transistors from the good side in, and the resistors went up in smoke again.
i noted the voltages are high on the pair of NPN transistors, most notably where it should be 7 volts, its about 30 or 40.
i cant get MPSA42s where i live, but the MJE340 looks pretty close, are these a suitable replacement?
also, is there a possibility i nuked the output transformers?
heres the schem for you, sorry perreaux.
 |
|
|
| Zero Cool |
WHAT your using a perreaux as a BASS AMP???? OMG that poor amp. such a nice piece of audiophilia being bashed about in road rack. Oh the horror. Thats like using the hood of a Porsche for a work bench or saw horse!
I had heard rumours of guys using Brystons for PA's and Bass rigs and even saw a photo of a rack of Brystons at the Sky dome in Toronto for the PA and that was shocking enough but oh my...Somebody please rescue that amp and give it the precious home it deserves!
I'm going to go cry now.. |
|
|
| Zero Cool |
On a more serious note. check for oscillations. the Hitachi output mosfets have a tendancy to go bad when drive hard. the gate capacitance goes crazy and the amp will start to oscillate causing the driver stage to draw LOTS of current.
Been though the same thing with Big Haflers that were used as subwoofer amplifiers in Theaters. they would beat the **** out of them and they wouldnt last long. |
|
|
| anatech |
Hi James,
Get the parts from Perreaux then. MJE340 might be a bit slow but you never know.
If the gate resistors are burning, the output FETs are shot. BTW, never swap parts from a working channel into a non-working channel unless you have an awful lot of experience. It would involve testing those parts first as well.
Hi Zero Cool,
I have seen many over heated racks of Bryston used for various things. I don't care for those amps but I do agree that they do not do as well in those applications as the Pro amps designed for that abuse. While I do question the use of a Perreaux as a bass amp, it is James' amp. I do expect those clamping zeners to conduct on a semi-regular bassis.
-Chris |
|
|
| jamesm88 |
righto back with more info.
im preeeetty sure the output FETS are buggered. i did an in circuit test and on the good side, all the FETS measure infinite resistance gate to drain and gate to source. on the blown side they measure infinite gate to source, but there is some low (about 400 ohms on some, about 100 on others) resistances on gate to drain. im fairly certain that means theyre stuffed.
so apparently you cant get these 2SK135 and 2SJ50 transistors any more. the best i can get is 2SK1058 and 2SJ162. i had a quick look at the data sheets, and they look pretty similar unless ive missed something, but ill leave the final decision up to you more experienced folk. if they arent any good to me, what can i use as a substitute without having to scrounge around for NOS items?
as for using it as a bass amp, well its a mates amp, and it sounds bloody good! |
|
|
| nickds1 |
The circuit as posted is very very similar to the Perreaux 2150B - a example of which I recently rebuilt - it had similar symptoms to yours.
I have the service manual for the 2150B which you may be able to get from Perreaux - the bias set up procedures are fairly complex, and you have to be pretty sure of yourself as it involves replacing the fuses with mA meters...
In my case, the PSLN/PSLP (MPSA42/92 as correctly mentioned by a previous poster) were ok, but the output FETs in one channel were defunct - in fact one actually had a blown hole in the side ;-)
These output FETs are prone to exploding - don;t connect a loadspeaker during setup (you'll kill the speaker) - build a dummy load with non-inductive resistors & a BIG heatsink... You can see mine in the photo below - the dummy load (blue connecting leads) is next to the amp on the far right... Nick's testbench
Perreaux recommend powering these amps up with a Variac when testing to limit the smoke & flames. You can make one simply by using a few (e.g. 3) mains light bulb in batten holders in series - when first switching on, have all bulbs in series with the mains, then if you are happy, short one out, then if happy, take the next one out etc. I've made this up on a board with some switches - it's saved more than one amp...
If you want to know the bias setup procedure, let me know.
HTH |
|
|
| anatech |
Hi nickds1,
Would you be at all surprised to hear I like your collection of equipment? ;) What kind of work do you do mostly?
-Chris |
|
|
| nickds1 |
| quote: | Originally posted by anatech
What kind of work do you do mostly? |
As little as possible ;-)
My degree is EE, but I've never done it professionally. I do odd stuff (maths/IT) for exchanges & hedge funds mainly...
I like nixies & used to build a lot of SS amps for friends, but I'm just getting into building tube amps... a whole new genre to learn about... I've been quietly (mostly) hanging about here for 6 months or so... |
|
|
| anatech |
Hi nickds1,
I'm still drooling over the pile 'o gear! Mostly RF or analog bandwidth?
-Chris |
|
|
| jamesm88 |
i think youd better give me a rundown on the bias procedures, just in case. i dont know if it was present on the amp you were working on, but in the 3000b the trimpots have little smears of paint on them so you can match up where they were if they get moved, so i figured id just put them back there.
as far as i know, you just take a fuse out and put your multimeter in set to mA and make sure they are consistent with the good side? and will i have to set the bias differently to suit the new output FETs?
thanks alot for your help guys, its saved my sanity for a little while longer |
|
|
| anatech |
Hi James,
Read the posts over again. Use resistors, not your meter mA scale! Just measure the voltage drop. If you have an inexpensive meter, use 1R0 resistors. try to contact the service department for the proper procedures.
Check the diodes from gate to source before taking the outputs out. If you have never replaced output transistors before, I would strongly recommend you take the amp to a good technician. Perreaux would be an excellent choice. There are so many things that can go wrong. Right down to screwing the outputs in too tight. Failures may take a long time to show up, or not.
Sometimes it really is cheaper to have certain things done professionally. BTW, I personally do not consider a TV technician (or car stereo) as competent. Some are, most are not.
-Chris |
|
|
| nickds1 |
| quote: | Originally posted by jamesm88
i think youd better give me a rundown on the bias procedures, just in case. i dont know if it was present on the amp you were working on, but in the 3000b the trimpots have little smears of paint on them so you can match up where they were if they get moved, so i figured id just put them back there.
as far as i know, you just take a fuse out and put your multimeter in set to mA and make sure they are consistent with the good side? and will i have to set the bias differently to suit the new output FETs?
thanks alot for your help guys, its saved my sanity for a little while longer |
Hope this helps (sorry it's big):
PMF 2150B bias setup |
|
|
| zeonrider |
Nice work.
Regards zeoN_Rider |
|
|
|
