The feedback ccs have very bad temperature stability because there is no mechanism to compensate the Q1 2.1mV/cº tempco, the led have near the same tempco from the bjt efectively compensate the led_ccs.
In the book I have his suggest this one,
the extra resistor match the tempco of the sense diodes that are 1,7mv/cº to 2,1mv/cº of the transistor, is simple, but very effective.
The predrivers have to be mounted in the heat-sink, if they left out of the heatsink there is no thermal feedback from them, and a increase in temperature inside the amplifier case will increase the bias current of the output transistors, making the distortion increase.
Exactly that one I used In my TT amp.
I adopted that one with two transistors bias spreader to allowed symmetric drive, and in the case that the predrivers are not mounted on the same heath sink with the drivers, one transistor from the bias spreader could be left not heathsinked and thus follows inside the amplifier case temperature and compensate the predrivers. If the predrivers are on the same heath sink with the drivers thy will heat more with no need. I have to simulate it yet. What do you think about it Sergio?
Damir, please dont get me wrong. I do believe that your circuit will work, but I have a tendency to follow the simplest soluction if that soluction is effective. Simpler soluctions tend to give less trouble
.I have spend a lot of time trying to find the best soluction. I will present it here , but is up to you the final word, after all is your project
.Good night friends.
Ok, i see it now . The thing that i didn't liked in your soluction is the diodes at the drivers emitters, your modification is a very good idea , but for better thermal stability is better to put the 2 bias sense transistors in the heatsink with the predriver and driver. I will use your bias spreader, but will discard the diodes in the drivers emmiters.
Sergio, I suggest that we use what we agree is better solution, this is more your project than mine, EC at least.
To use TT diodes (anyhow they are here unused in TT transistors, but this solution complicate the layout a bit) in the drivers emitters is Cordell idea (showed in Thermal Trak thread some time ago) and in this case you don't need a capacitor parallel to the emitters resistor.
Valery, input pot is there if you want to use this amp without preamplifier, so it is up to you.
For P2 you can put provision for both, the trimmer pot and a resistor in series with the pot and decide later during the test if we need it.
What do you think is the best place to connect the DC servo?
br Damir
OK, good, I will reserve the plug for P4 and do both options for P2.
With regards to the servo, I would use a straightforward approach and connect it to the input via some 100K resistor - small enough influence on input impedance and enough current to control DC offset...
Good day RNMarch.
The internal speed of the power amplifier is 400 V/us . The SR is limited by the input filter.
You can easily get 200w-250w/8 if use the amplifier in bridge configuration, use a good quality transformer to get the differential signal.
Last edited:
I have test your bias-spreader today and it worked very well, 2 diodes is enough to thermal compensate the output stage. using the 4 diodes can cause overcompensation.
It is only need to use two thermaltrak transistors (NJL3281D/NJL1302D) the other pair can be a (MJL3281A/MJL1302A) as it is the same transistor, but without the sense diode, and it costs half the price. of-course it can be use 4 thermaltrak transistors but one pair will have the diodes without use.
this is a image of the bias spreader. R1,R3,R5 have to be static resistors, R2 and R5 can be in serie with trim-pots to set the output Bias current.
With the values in the picture the tempco is 4,35mv/cº for compensation of the output transistors.
The two transistors will have to share a heat-sink with the prédriver and driver transistor for better thermal compensation.
Attachments
Last edited:
Thank you Hugh,
Here is the amp with better CCS suggested by Sergio. I put it only in use in the bias spreader, I think it's not needed in the EC part, what do you think Sergio? As this CCS eats more voltage I increased second power supply voltage to get a room for better clipping.
As Sergio does not like to use second pair of the TT diodes, they can be replaced with a resistor parallel with a capacitor, and other output pair with ordinary power transistors of the same characteristics (by the way I plaid with that idea in my TT amp here http://www.diyaudio.com/forums/solid-state/216780-tt-amp-200w-8ohm-701w-2ohm.html#post3104438).
BR Damir
Attachments
Last edited:
This is the circuit I have been testing, I will start to make a test pcb with it for evoluate the concept and test other ideas that I have.
This circuit as is, does not have any protection circuit yet
.
R31/28 are Trimming potentiometers for setting the bias, r32 is a Trimming potentiometers for fine calibration of the distortion cancellation.
The decoupling caps should be close to the respective transistors especially those 220nf ones that should also have low Inductance (ESL).
what would you like to be include or modified ?
PS: spotted already an error R13 and R26 need to have lower value (1K)
This circuit as is, does not have any protection circuit yet
.R31/28 are Trimming potentiometers for setting the bias, r32 is a Trimming potentiometers for fine calibration of the distortion cancellation.
The decoupling caps should be close to the respective transistors especially those 220nf ones that should also have low Inductance (ESL).
what would you like to be include or modified ?
PS: spotted already an error
R13 and R26 need to have lower value (1K)Attachments
Last edited:
Excuse me for coming in late here.
Doesn't this design error correct only for an ideal resistive 8 ohm load?
Isn't the idea of error correction to correct for the 'actual' load, ie; a complex
speaker load.
It just doesn't seem like something that will work very well beyond the
simulator - which we all appear to be obsessed with these days
cheers
Terry
Excuse me for coming in late here.
Doesn't this design error correct only for an ideal resistive 8 ohm load?
Isn't the idea of error correction to correct for the 'actual' load, ie; a complex
speaker load.
It just doesn't seem like something that will work very well beyond the
simulator - which we all appear to be obsessed with these days
cheers
Terry
are you serious ???Terry not get me wrong ... but I wonder how long you lost looking for the scheme to reach that conclusion, surely not much.
This is a slightly different concept of error correction than normally found. so the faster and right start is to simulate and then build. of course that works with any load, and not only with 8 ohms resistive.
any way... welcome
Last edited:
This is the circuit I have been testing, I will start to make a test pcb with it for evoluate the concept and test other ideas that I have.
This circuit as is, does not have any protection circuit yet
R31/28 are Trimming potentiometers for setting the bias, r32 is a Trimming potentiometers for fine calibration of the distortion cancellation.
The decoupling caps should be close to the respective transistors especially those 220nf ones that should also have low Inductance (ESL).
what would you like to be include or modified ?
PS: spotted already an error
Hi Sergio,
I'd like to discuss one point, related to bias setting approach.
As the spreader here is not "2-sided" (in this case it would not matter if it is symmetric or not), but 3-sided (with central connection on the left and requirement for its shoulders - top and bottom - being balanced), it will be very difficult (if not impossible) to adjust the bias and keep the spreader balanced at the same time with 2 trimmers (R31/28) as shown. See what I mean?
I would leave one of this trimmers for balancing the shoulders, and put the second one in series with R14 for adjusting the bias symmetrically. I understand - adjusting R14 also influences multiplication factor, but adjustment required (and though influence on multiplication) is going to be rather low.
Does it make sense?
Cheers,
Valery
Good morning Valery.
I think that a few mv of difference between the upper part from the lower part do not make any difference as the servo will compensate that. It only make sense to use two trimmers in a test pcb. Then one see what Is the value of the needed resistor, in a final pcb only one trimmer is need for bias adjust , R31 can be a fixed resistor. I think that is also better to leave the multiplication factor fixed.
Do you agree?
I think that a few mv of difference between the upper part from the lower part do not make any difference as the servo will compensate that. It only make sense to use two trimmers in a test pcb. Then one see what Is the value of the needed resistor, in a final pcb only one trimmer is need for bias adjust , R31 can be a fixed resistor. I think that is also better to leave the multiplication factor fixed.
Do you agree?
I'm just not sure how critical it is for error correction circuit, if the spreader shoulders are slightly unbalanced. If not critical - then no problem, I agree.