Hey PMA....Where's the Bootstrapping in MikeB's amp.....PMA said:Oh, Lord. Quasi-complementary + bootstrap. You missed the thread.
MikeB said:
No , Mike its isn't like that...the circuit you have pointed to lacks thermal compensation alot......therefore Thermal runaway kisses that amp every Ten minutes.....😀 😀 😉
Workhorse said:
Nowhere, but as there was an idea of quasi-complementary output I thought it would not be far from recommendation to bootstraping 😉 . Both are very popular in DIYaudio.
I had made quite a lot of amps with both of these solutions and I did not like neither quasi-complement., nor bootstrap.
PMA said:
Hi Pavel,
I believe that I have all of the jumpers in but I'll double check when I get home from work today.
The trim pots are not having much of an affect right now. Probably part of the problem.
Thanks, Terry
Terry,
look harder for shorts, there is one between the neg and ground in the VAS area. The other board has a similar problem but the second diff is likely fried.
Look, what is with tinning the pcb?
You got lotsa nice components there, I strongly suggest you spend a couple of hours with the pcb express software and remake you board in a nicer way. Last time I checked you could get 3 prototypes made for 60 bucks and they come professionally made an pre tinned.
Unsolder all the components check them (everything including passives) and, if they are not overheated by tinning, resolder on the new boards.
look harder for shorts, there is one between the neg and ground in the VAS area. The other board has a similar problem but the second diff is likely fried.
Look, what is with tinning the pcb?
You got lotsa nice components there, I strongly suggest you spend a couple of hours with the pcb express software and remake you board in a nicer way. Last time I checked you could get 3 prototypes made for 60 bucks and they come professionally made an pre tinned.
Unsolder all the components check them (everything including passives) and, if they are not overheated by tinning, resolder on the new boards.
I am using 180 ohms in the rails...and i had made comparison with the 22 standard
rail resistors and the 180 that i have used.
I had lost , something around 2 Volts Peak to Peak in the power output, but sounded better with bigger resistor.
I think 22 ohms is something near "nothing" to avoid voltage fluctuations to the input circuits under the moments that supply will be hard drained, and voltage will reduce, as a consequence, as you know...reducing the input circuits voltage too.....so...not protecting nothing.... when used normally to stabilize the input circuits voltage... that small resistor will work perfect to Simulators, because simulators do not loose voltage when amplifier drain current...hehe
Real world produces big voltage variation in the supply rails when some strong bass drain supply current.... to separate power circuits from input circuits, better use bigger resistors.. and big condensers too...bypassed with 100N
I observed someone saying to use 3 ohms.... it could be some mistake...the guy forgot one or two zero digits after.... well, this is what i suppose, because the one is a skilled guy.
Well...simulator electronics are interesting...supply voltage is stable....but real world...supply is not stable.... voltage is always fluctuating if you are not using an electronic stabilized fast response supply.
With 22 ohms..... around 500 milivolts stays over the resistor (if i remember the correct voltage there)..... so.... 36 volts to output and 35.5 volts to input...... but when you drive full power...your supply will reduce to something around 32 V.....(If a normal, not enormous supply)..... and your input goes to 31.5 that you will measure on the rails.
If you have bigger loss of voltage in those rail resistors...let's see something very big as 5 volts (for instance).... the voltage variations will not be perceived in the input circuits since you have used 180 ohms or bigger resistor in series with the rails, creating a second reduced voltage to feed the input stages, and they are more sensitive to those voltage variations...so....need to be stable, with good capacitor or circuits to stabilize there..using noise zener or some stabilize transistor.
Normally, after the rail resistor exists an electrolitic condenser, that one, beeing charged, will keep voltage stable or will allow a slow reduce of voltage.
Your loss of output power, will be compensated by much more quality, in special when driving the amplifier hard.... and you can increase your supply voltage, to force the input circuit to work with the voltage used to design the amplifier...the correct voltage.
Because of that, one amplifier that i am using is working with 42 volts simetrical, and loosing 6 volts in that resistor...so...input is working with standard voltages (...and output will fluctuate, but will not go lower than the voltage Michael used to Simulate...so, dinamically, and hardly used, it will work the way Michael designed it.
regards,
Carlos
rail resistors and the 180 that i have used.
I had lost , something around 2 Volts Peak to Peak in the power output, but sounded better with bigger resistor.
I think 22 ohms is something near "nothing" to avoid voltage fluctuations to the input circuits under the moments that supply will be hard drained, and voltage will reduce, as a consequence, as you know...reducing the input circuits voltage too.....so...not protecting nothing.... when used normally to stabilize the input circuits voltage... that small resistor will work perfect to Simulators, because simulators do not loose voltage when amplifier drain current...hehe
Real world produces big voltage variation in the supply rails when some strong bass drain supply current.... to separate power circuits from input circuits, better use bigger resistors.. and big condensers too...bypassed with 100N
I observed someone saying to use 3 ohms.... it could be some mistake...the guy forgot one or two zero digits after.... well, this is what i suppose, because the one is a skilled guy.
Well...simulator electronics are interesting...supply voltage is stable....but real world...supply is not stable.... voltage is always fluctuating if you are not using an electronic stabilized fast response supply.
With 22 ohms..... around 500 milivolts stays over the resistor (if i remember the correct voltage there)..... so.... 36 volts to output and 35.5 volts to input...... but when you drive full power...your supply will reduce to something around 32 V.....(If a normal, not enormous supply)..... and your input goes to 31.5 that you will measure on the rails.
If you have bigger loss of voltage in those rail resistors...let's see something very big as 5 volts (for instance).... the voltage variations will not be perceived in the input circuits since you have used 180 ohms or bigger resistor in series with the rails, creating a second reduced voltage to feed the input stages, and they are more sensitive to those voltage variations...so....need to be stable, with good capacitor or circuits to stabilize there..using noise zener or some stabilize transistor.
Normally, after the rail resistor exists an electrolitic condenser, that one, beeing charged, will keep voltage stable or will allow a slow reduce of voltage.
Your loss of output power, will be compensated by much more quality, in special when driving the amplifier hard.... and you can increase your supply voltage, to force the input circuit to work with the voltage used to design the amplifier...the correct voltage.
Because of that, one amplifier that i am using is working with 42 volts simetrical, and loosing 6 volts in that resistor...so...input is working with standard voltages (...and output will fluctuate, but will not go lower than the voltage Michael used to Simulate...so, dinamically, and hardly used, it will work the way Michael designed it.
regards,
Carlos
"that small resistor will work perfect to Simulators, because simulators do not loose voltage when amplifier drain current...hehe"
Hmm, Carlos, i did weaken the supply in sims, otherwise the 22ohms would not have had any effect... Without the elyt following the rail resistor, it would be of course worthless !
Mike
Hmm, Carlos, i did weaken the supply in sims, otherwise the 22ohms would not have had any effect... Without the elyt following the rail resistor, it would be of course worthless !
Mike
Have you tried bigger ones?
Real world sounds better...to years.
Off course including the needed electrolitic condensers, because without of them are non sense.
Let's say 220 ohms...increasing the supply value not to get out from your designed voltage.
And please, let me know if you simulator has that capacity, to reduce the supply voltage in accordance with the drain current.
Well, your simulator may be "a must"....i never heard that could work that way..congratulations.
Have you designed your simulator Michael?
Have you read my text...is it confused?....i think it is clear and correct.
regards,
Carlos
Real world sounds better...to years.
Off course including the needed electrolitic condensers, because without of them are non sense.
Let's say 220 ohms...increasing the supply value not to get out from your designed voltage.
And please, let me know if you simulator has that capacity, to reduce the supply voltage in accordance with the drain current.
Well, your simulator may be "a must"....i never heard that could work that way..congratulations.
Have you designed your simulator Michael?
Have you read my text...is it confused?....i think it is clear and correct.
regards,
Carlos
MikeB said:
Hi Mike,
Thanks for taking the time to help me with this.
1) None of the pins are grounded. I didn't clean the board after soldering, I will do that.
2) I used all of the transistors that are listed on the BOM on your website.
3) C14 is 22pf like the BOM states. They are large because of the voltage. I probably should have found some that were lower voltage. 🙁
If you think 10pf works better, why not change the BOM?
4) I just stuck that electro in there so I could test it. The cap I ordered was wrong.
I realized R7 need the coil too late. I will change it. The leads are too short now to use. I will have to get more. I think I need a larger diameter resistor for the coil anyway.
The bias pots don't seem to be having any affect. Could be the problem is around them.
Blessings, Terry
Re: I am using 180 ohms in the rails...and i had made comparison with the 22 standard
Carlos, I would strongly request for you not to post those nonsenses.
I spoke about 3R resistors in supply rails as a current protection. When I first start the amp, I put them before PCB supply pins +V and -V (before fuses). This reduces max. current not to burn output device immediately.
They have nothing to do with 22R resistors that are used for filtering of input stage.
destroyer X said:
Carlos, I would strongly request for you not to post those nonsenses.
I spoke about 3R resistors in supply rails as a current protection. When I first start the amp, I put them before PCB supply pins +V and -V (before fuses). This reduces max. current not to burn output device immediately.
They have nothing to do with 22R resistors that are used for filtering of input stage.
Hi Terry, i now recommend the following:
- Measure the resistors and all transistors, replace the defect ones. Some of them will be fried.
- Clean the solderside of the board and search carefully for solder bridges and cold solders, don't hurry.
- place 100ohms instead of the fuses
- when testing / powering up, set the variac to minimum voltage, then slowly increase the voltage and measure the voltage across the 100ohms. If this voltage goes much above 1v, immediately turn off and continue searching.
- Make that with one board after another
The bias pots will start to work only if amplifier is functional, i think your problem is around the VAS, in the middle with the 5 bjts, but search everywhere.
For the case you make new boards, skip the tinning... Tinning is only good if done galvanically.
You're right, i should have updated my page regarding the 22pF, sorry...
Mike
- Measure the resistors and all transistors, replace the defect ones. Some of them will be fried.
- Clean the solderside of the board and search carefully for solder bridges and cold solders, don't hurry.
- place 100ohms instead of the fuses
- when testing / powering up, set the variac to minimum voltage, then slowly increase the voltage and measure the voltage across the 100ohms. If this voltage goes much above 1v, immediately turn off and continue searching.
- Make that with one board after another
The bias pots will start to work only if amplifier is functional, i think your problem is around the VAS, in the middle with the 5 bjts, but search everywhere.
For the case you make new boards, skip the tinning... Tinning is only good if done galvanically.
You're right, i should have updated my page regarding the 22pF, sorry...
Mike
.... the other possibility is that I still have 2 perfect profi double sided PCBs with solder mask .... 😉
I can sell these 2 pcs, but have not got intention to start mass production.
15.00 EUR per piece + 5.00 EUR shipment within Europe. Send me an e-mail, please.
15.00 EUR per piece + 5.00 EUR shipment within Europe. Send me an e-mail, please.
Yes Pavel, was my mistake...i was talking about the 22R resistor, not this one
I am sorry, my mistake.
The 22 ohms is the one that is too much small...it is inside amplifier, not external as protection, this one is to reduce voltage to input stages, to avoid that voltage variations disturb the first stages...normally used between driver and voltage amplifier transistors...you know...the traditional one.
This 22 ohms, to separate the input circuits related to output voltage variations use normally an electrolitic condenser, 47uF or 100uF normaly following the internal circuit rail resistors.
Those ones to limit current when board may have problems, or even to help the bias adjustment is another story, and value will depend to the output transistor current limit.
Both of them are very basic, and i am surprised that Michael is beeing using so low value, when was already informed that real tests were made, and 22 ohms was clearly not adequated...beeing many times smaller than the needs.
Exception is some huge power supply, that will variate nothing draining 6 amperes plus 6 amperes.....both channels driving hard with 4 ohms and 37 volts keept stable...using power regulator, stabilize power supply or something that keep the supply constant....my simulator keep the supply constant..you can drain 1 ampere of a million amperes and voltage do not fall a milivolt..ahahaha...very unreal!
regards,
Carlos
I am sorry, my mistake.
The 22 ohms is the one that is too much small...it is inside amplifier, not external as protection, this one is to reduce voltage to input stages, to avoid that voltage variations disturb the first stages...normally used between driver and voltage amplifier transistors...you know...the traditional one.
This 22 ohms, to separate the input circuits related to output voltage variations use normally an electrolitic condenser, 47uF or 100uF normaly following the internal circuit rail resistors.
Those ones to limit current when board may have problems, or even to help the bias adjustment is another story, and value will depend to the output transistor current limit.
Both of them are very basic, and i am surprised that Michael is beeing using so low value, when was already informed that real tests were made, and 22 ohms was clearly not adequated...beeing many times smaller than the needs.
Exception is some huge power supply, that will variate nothing draining 6 amperes plus 6 amperes.....both channels driving hard with 4 ohms and 37 volts keept stable...using power regulator, stabilize power supply or something that keep the supply constant....my simulator keep the supply constant..you can drain 1 ampere of a million amperes and voltage do not fall a milivolt..ahahaha...very unreal!
regards,
Carlos
Hi Carlos !
I have a very high end toroid (250VA), not the stuff you buy around the corner...
In sims, it is VERY easy to simulate weak powersupply, add 1ohm resistor in series with the powersupply, this gives you realistic poor power... Drops 1v per ampere sucked.
22 ohms were way enough to kill the distortions created by such weak powersupply. Higher values created new problems...
Mike
I have a very high end toroid (250VA), not the stuff you buy around the corner...
In sims, it is VERY easy to simulate weak powersupply, add 1ohm resistor in series with the powersupply, this gives you realistic poor power... Drops 1v per ampere sucked.
22 ohms were way enough to kill the distortions created by such weak powersupply. Higher values created new problems...
Mike
Even with 2 transformers, one to each channel, with good filter condensers
inject audio and adjust volume till the amplifier start to distort, using 4 ohms speakers.
Now use a Voltimeter, those that have needle, the analogic ones.
Measure voltage after the 22 ohms resistors.... referenced with ground....measure positive and negative rail.
Can you see variations?...the needle is moving downwards.... this is what i am telling you...real music...real non street corner transformers and real good electrolitic condensers.
All those...with unreal 22 ohms resistor, will produce real variation of voltages.
If some problem to measure...well, ask John, he may help you.
Audio to inject is normal audio program, from CD for instance, and the distortion i am telling you, this threshold, is something your ears will perceive, and may corresponds with some start of clipping in waveform.
Just to remember you....use DC voltage scale...ahahahaha!
Using those yellow chinese multimeters, those we find in the corners, the voltage will go variating.... and the numbers will not stop to change.... if you have not a "hold" function or "average" function, use those old multimeters that have an oscilating needle that will dance following the variations of voltage.
regards,
Carlos
inject audio and adjust volume till the amplifier start to distort, using 4 ohms speakers.
Now use a Voltimeter, those that have needle, the analogic ones.
Measure voltage after the 22 ohms resistors.... referenced with ground....measure positive and negative rail.
Can you see variations?...the needle is moving downwards.... this is what i am telling you...real music...real non street corner transformers and real good electrolitic condensers.
All those...with unreal 22 ohms resistor, will produce real variation of voltages.
If some problem to measure...well, ask John, he may help you.
Audio to inject is normal audio program, from CD for instance, and the distortion i am telling you, this threshold, is something your ears will perceive, and may corresponds with some start of clipping in waveform.
Just to remember you....use DC voltage scale...ahahahaha!
Using those yellow chinese multimeters, those we find in the corners, the voltage will go variating.... and the numbers will not stop to change.... if you have not a "hold" function or "average" function, use those old multimeters that have an oscilating needle that will dance following the variations of voltage.
regards,
Carlos
Well - regarding this 22R story - it is all about time constant, in Mike's case 22 ohm x 100 uF, which is 2.2 ms (miliseconds). This is really not too much .😉, and probably might be 10 times more at least. There are 2 possibilities, to increase 22R resistor or to increase 100uF filtering cap. The later (increasing the cap) has one disadvantage, i.e. larger hum current injection into PCB. A question of check in a real circuit.
Carlos, thank you, yes, this is the purpose of 3R res protection for the 1st startup. My supply voltage is +/- 27V, so it would limit current to +/-9A max. when serious failure occured.
Carlos, thank you, yes, this is the purpose of 3R res protection for the 1st startup. My supply voltage is +/- 27V, so it would limit current to +/-9A max. when serious failure occured.
Terry,
From the pictures you have posted, it is quite evident that the Emitter and Collector of MJW1302A are shorted. This must be one of the problems, if not the only.
From the pictures you have posted, it is quite evident that the Emitter and Collector of MJW1302A are shorted. This must be one of the problems, if not the only.