Amplifier Testing

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
THE FIT THROUGH SCREW

hello andrew t thanks for posting .....

the truth is that the screw its shelf that hold the heatsink created the mess ......

but at the time no miller capacitor existed between B-C of vas amplifier MJE 340.....

miller cap was recomended from bora later on

the 50v dc was coming from gennearlly instability of the amp cause of local ocilation in the vas stage


removed the screw lift up MJE 340 and used "onatop"hetasink away from board and added caps of 33pf at 56+56 vdc boards and rise up to 68pf at boards that operate with almost 75+75vdc....


thanks for your recomendation about power of your amp but i am hooked to this legend since i allready have about 30 boards and almost 200 mosfets of IR IRFP250....

also except amps i construct from scratch i had to refab some that psu 66+66 allready existed together with box caps and all the others ....the original for example inter m 2300 was totally burned and pcb was in mars ....totaly no point of repair so i choosed to pute legend inside

real life stories and mods .....
 
lets clear things

i am sakis petropoulos and yes amplifier legend 4 and 8 was given to me from dr Bora for free and also given as request to operate in voltage from 56 to 66 vdc per rail

well this amplifier is working perfectly in voltage 40+40 volts but no matter how much i tried so far to operate at 66+66 volts i end up with burned mosfets burned speakers and a lot of smoke

i have grate respect for dr BORA and also his designes i am extremelly thankfull cause he gave this for free to me and also willing to help and suport me with alot of suport
enenthough we all very well know that he is in a difficult moment of his life ......

to my understanding maybe dr Bra missed that my rails are 66+66 volts and may be many things need to be chanched to operate at this voltage ....thats ok he done a lot so far and i know very well that may be he canot do more at the time so i asked help from all you forum guys and seems that after a long time i am getting somewhere .....


also didnt know that dr Bora was sick so i made posts saying that probably dr Bora didnt help me out i could neverthink that he might be in a hospital

I APOLOGIZE FOR THAT ......

but otherwise i never altered or chanched any of my request or the amplifier spece

sakis petropoulos
 
ok, you mounted the heat sink flat on the board on the other side where the screw is, the added capacitance is most likely from the heat sink to the two traces on the board that go past the screw, since they are under, and parallel to the heat sink. also be aware that if you change the rail voltages, the junction capacitances of the bipolar transistors change. the b-c junction of a bipolar transistor, when reversed biased acts as a varactor, and the higher the reverse bias, the lower the capacitance. so you may need to increase the compensation capacitance by about 10pf to make up the difference.

as far as going from +/-40V to +/-66V rails, understand that now, all of the output devices, drivers, predrivers, and anything else running from the rails, needs to be rated at much higher voltages. Vceo of the output devices needs to be at least 150V, if not a bit higher, since the difference between the rails is now 132V, not the 80V that it was before. i would actually go with 250V devices if possible, just because you are driving primarily inductive loads, and you need as large a SOA as you can get.
 
thanks for the advice

but as i said befor and almost a gozillion times i DIDNT CHANGE TO HIGHER RAILS VOLTAGE !!!!!the amp was given to me to operate at 66+66 volts and it doesnt ...

i still belive that this was just a missunderstanding or an overestimate of bora the point is the same that i am now stucked with 150 burned mosfets ...another 250 brand new ....about 30 boards ....and simply dont know what to do with them ...

the screw was a mistake but even removing this or adding vas capacitor didnt solve much

so????? any ideas?????

THANKS FOR POSTING !!! sakis .....
 
Hi Sakis,
the amp schematic shows a supply voltage range of +-40 to +-60Vdc.
The amp design has used resistor feeds for all the current sources/sinks.
Each of the resistor sources/sinks will carry different currents as the supply voltage changes.
The operating points and the openloop amplification (gain) will vary with supply voltage.
The feedback factor will change as the open loop gain varies.
The stability of the amplifier will change as the supply voltage varies.

I would be very surprised if this amplifier would operate correctly over the full range of 40 to 60 volts. I think some component values will need to be adjusted to work at different supply voltages within the recommended range.

You have confirmed that your version did not blow up on +-40Vdc.
You have confirmed that your version repeatedly blows up on +-66Vdc.

Can you see that this schematic and the increased supply voltage is the cause of the reliability problem?
 
o god !!!!!

it seems that you dont understand much andrew t

folowing your posts through the forum about my problema and also other amps, of other people it seems that you have almost a bad opinion for any amp existing in the forum.... i ve seen your posts to many others that construct much more advanced amps than the one i try !!!!!

YES !!!! i confirm that amp looks stable at 40 volts
YES!!!!! i confirm that the schematic say 40-60 volts

do you think that 66+66 volts is the problem ???? well i tryied to operate at 55+55 and still the same ...very unstable

the problem is that probably this amp is designed to run tops 45+45 volts

fianlly changing the vas curent........and the output resistors to 0.27R/5W amp is working but how well i dont know at leasts its steady and sounds good

you see andrew t except people that drive a Rolls might be others that drive a hundai

i f you wish to call this amp a hundai is fine with me

i am gona stick to it

thanks for posting andrew t any way.... and eventhough even in this post or the private contact we had i ever manged to come up with any usefull results ......
 
another small thing

many people so far gave there opinions about this amp some of them good and some of them bad .....but finally thats not the issue .....

meaning that yet nobody come up with some results saying change this change that ...simple things and will work ..... critisism is fine but results ....simply nor exist

since i make my shelf clear..... first i have no design scills !!! then my request originally was for a circuit that operates at 42+42 ac cause existing boxes i have cannot change ...and the boxes are almost 15

and then the power of this is ok to be around 150w at 8R but in a SOA...

at this point i produce 187.5 w rms @ 8R with legend 4 and almost 225w with legend 8.... sound is ok power is way too much and more than enough and also looks stable

probably some small changes here and there will make almost perfect

andrew t said to me that this design will never make 800khz ....and yes i will write that once more 800khz !!!!!!

one of us is missing the point here !!!! 800khz is am transmiter not audio amp !!!!! i will be so happy with 25khz

but on th other hand thats me .....
 
Hi Sakis,

I saw your posts about this amplifier, and note that there appears to be a specific problem related to instability as you increase the rail voltage. (Not unusual.)

I note that TR1 base has a series 1k input resistor. This increases the impedance of the NFB sensing reference point.

You could try shorting it out. If this upsets offset, make the 10k bias resistor into 11k.

Also the 'mica' capacitor betwee VAS and TR2 base might be better with a series 1k resistor.

Sometimes a resistor of 47k to 100k between VAS collector and ground improves stability too.

Good Luck ......... Graham.
 
Hi Sakis,

Having a lot of respect for Carlos I accept his recommendation of you and respect you for that. But I think it also needs to be recognised that giving the advice you need in this way has a lot of difficulties. You say that you are not too technical - accepted. But I sense that here is a problem where perhaps simple things might be overlooked because we are not there with you. I would have started certain measurements (checks) long ago, but it is impossible to sit thousands of km away and guess right.

That is why you have frustration and we cannot seem to help. At a stage in the heat sink argument I was going to ask whether the heat sink was ever earthed (with isolation of the transistor), but then that did not appear to be the problem.

Regarding the higher voltage problem all we know here is that you get no problems with 40V but you do with 66V. The logical conclusion must then be that something is changing - whether the circuit was designed for 66V or even higher is unimportant. You have a problem - that is important.

I would feel (and it is not good news to you) that, frankly, it will be luck if we can suggest something that would help. It probably needs a careful on-board investigation with scope and signal generator, especially as you are destroying expensive components. E.g. I would put in a square wave (low amplitude) and watch carefully with a scope on the output until the first sign of oscillation as the h.t. is slowly turned up with a variac. I would also have current meters in the power leads and watch them too, ready to switch off at any time. But you can perhaps sense the difficulty. How do I tell you what then to do without being present to recognise any symptoms? There are too many options.

I have been into transistor amplifiers since transistors have been invented and know this problem, and the difficulties. I have blown my share of transistors, but also saved some because I recognised symptoms and was able to correct. But there is no way I can put this on paper, even to an engineer - it would require many pages.

I sincerely hope that you can get someone to come to you and measure before you suffer more damage. I regret that I am unable to help, much as I would like to.

Regards
 
Well..i have to inform my friends that i am removing my recommendation about Mr Sakis

We have not a good agreement about Bora's chapter...related Bora's relationship together Sakis.

I do not think that Sakis was intirelly fair...even if Bora had failures, he could have a chance not to see his name published in the forum the way it was.

Related that, Sakis was kind, posting his apologize.

But something more were not entirelly fixed between me and Mr. Sakis Petropoulos, and this made us finish with our conversations.

I told Mr. Sakis that i do not want to continue those discussions, as when we put our hands in bad things, normally start to smell bad...also i told him that unless he enter the forum to bomb Bora, that i would make silence related his posts.

Sakis is now a forum coleague, not more my personal friend and have not my recomendation anymore.

regards,

Carlos
 
Well then, personally I am very sorry to hear that.

Still, the man had a problem with which all would like to help and some made good suggestions. But in general I hope folks would understand that in giving such advice one hopes to be lucky. I am not posting on behalf of others, but unless one is in possession of the same amplifier (and especially for semiconductor circuits), I feel confident that more members would agree with the difficulty of being able to help successfully.

Regards
 
Hi Johan,

Yes very difficult from afar, and unfortunately even seeing the circuit diagram does not give all the clues that are necessary. Long experienced constructors can often spot (hear) problems when an amplifier is directly in front of them.

Hi Sakis.

The input filter of the circuit you posted is 16 (sixteen) dB down at 800kHz.
Are you bypassing that filter, because although there are many amplifiers capable of amplifying at higher frequencies this is not normally claimed as if a working bandwidth characteristic ?

I've also just noticed that this one of those amplifier circuits having different 'grounds'. Nice in theory, but not always in practice.

Your output Zobel network, as with a heatsink ground, should go to INPUT (signal) ground, NOT a PSU ground. Also when you parallel output devices this is like having two output stages working together, so it might be worthwhile halving the resistance value of your Zobel.

Also note that all your 0.18 resistors plus the Zobel resistor ought not be plain wirewound, but non-inductive types.

Additionally, once an amplifier has 'blown' you need to check both the Zobel resistor and capacitor, for these can fail and allow the amplifier to blow up all over again as soon as new parts are fitted; the capacitor needs to be at least 250V rating as well.

Any photos so that we can better 'look over your shoulder' ?

You could also fit resistors in series with your fuses. Say 2.2 ohms. This would allow the fuse to blow before an output device fails !!!



Cheers ........ Graham.
 
complicated design vs simple design

Now finally perhaps you can understand that complicated - as you believe – designs, do not become without aim. An amplifier with constant current sources in all stages and in particular with transistors and not diodes - because these don’t have the beta factor - and current mirrors, and buffer (predriver – main driver ) stages it is positive under all circumstances and not changed the static or dynamic currents with the increasing or decreasing of supply voltage. These remain stable, and the only that we have to do for increasing the power is the addition of more transistors in parallel at output and the change from lower Vce transistors to higher Vce transistors. Also a VI limiter it is a transparent guard of the output devices as it acts only in heavy or short loading of output. Sorry our friends of Mosfets, but the BJTs in such circumstances they are proved more stable and it does not tease they are not so much fast – as you believe at least – from precious Mosfets. Believe that we haven't tried also Mosfets?
 
Fotios,

Also from me thanks for reminding about Zobels (I did not see the rest of that thread). Yes, very important. Although I also permanently load e.g. an 8 ohm output with 680 - 800 ohm to give resistive h.f. load (the power consumed is insignificant), one needs to remember that when it does "operate", it has the same power behind it that the loudspeaker would have.

Regards
 
ok guys ......

very nice of so many people at least willing to help .....

as about carlos if you no longer consider me as your friend ...i dont mind ...... personally i dont blame you or dr bora

it is obvious for one more time that one of us made a mistake consisdering voltage rails .......would you like me to take this blame ....?????? i can take it if you want and apologize to dr bora once more saying that this was my mistake ....will that make you happier ????? if its then i will do so ....

but still this is not solving the problem ( if ever possible to be solved )

as about fotios i am with him 100% but all these nice curent mirrors and contoled curent sources are perfect but my amp is not this kind of amplifier...its a simple thing and what ever comes with it


what i ask for is possible ..... the only thing you have to do is work litle with it an everything is going to be just fine ....for example rod elliots esp 101 project is almost the same thing but not quasi ....so its close..... litle things are to be made more to make if not perfect but shure better

as about johan potqieter and graham maynard wish full and polite people there is alot of material like pictures ,updated schematics measurments and many other info that i performed so far ....at your disosal if you are willing to work with me on this one.....

you can have all this info ...if and when you have the time to look at it .....

all you need to do is ask me and i will more than happy to provide


thank you all very much i really apriciate the will of all writing to me

thanks sakis
east_electronics@yahoo.gr
 
another thing

like my carlos said its very easy from such a distance suggestions or opinions may end up in smoke .....

i expect that also...... but the will to try and the effort of so many people will eventually bring results .....

on the other hand i am going to be the one who judge suggestions so belive me i am going to be very carefull before applying any mods .....

thanks sakis
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.