rawbas, the 2SC2073 is way too slow for the VAS transistor at an Ft of 4MHz. It may be OK with its complement 2SA940 as a driver but there are much better parts. Look at Greg Erskine's transistor types in Mayan's #114. He uses a CRT television driver too but BF469 is many times faster.
There are already several proposals similar to yours here, but always note the voltage limits imposed by the Vceo of the transistors they use. In the Digi-125, all the transistors are on their voltage limits but changing the types to bigger transistors affects the performance in other ways too.
That is your choice, but it may also mean the difference between an amplifier that just works OK and one that also sounds great. If possible, use BD139/40 if you are going for common parts but look for those with high Ft like old Philips or Unisonic, Fairchild versions. BD139-16 is also a good option, if available but note that ST Micro types are unspecified - could be anything.
There are already several proposals similar to yours here, but always note the voltage limits imposed by the Vceo of the transistors they use. In the Digi-125, all the transistors are on their voltage limits but changing the types to bigger transistors affects the performance in other ways too.
That is your choice, but it may also mean the difference between an amplifier that just works OK and one that also sounds great. If possible, use BD139/40 if you are going for common parts but look for those with high Ft like old Philips or Unisonic, Fairchild versions. BD139-16 is also a good option, if available but note that ST Micro types are unspecified - could be anything.
What do you mean by -+28VDC?. If this is DC, it is either -28V or +28V . You can't have both, unless you refer to 2 or more amplifiers, or some other part of the amplifier than the output.
Let's be certain of which particular schematic you have followed and what components you have decided to use first, otherwise we have to guess what you have done and that means poor suggestions because they are based on something different. If you have made your own PCB to suit the components, I guess there will be no problems if you checked the pinouts for your layout. However, if you copied the layout from someone else's, don't forget that pinouts for TO220 drivers are opposite to TO126 and of course pinouts for Japanese T092 style transistors are quite different to US and European (BC,BD etc). Google the pinouts if you are uncertain or have no data on your components.
If the VAS transistor or drivers are reversed from what they should be, either stage will be operating incorrectly or is now damaged. Check the voltage from base to emitter (Vbe) on all transistors and they should all be ~ 0.65V if working. Check also for shorts and other wrong connections.
You can be certain that the problems will be from errors in the PCB, errors in wiring or soldering or from using incorrect components. There's not much else that can be wrong.
Let's be certain of which particular schematic you have followed and what components you have decided to use first, otherwise we have to guess what you have done and that means poor suggestions because they are based on something different. If you have made your own PCB to suit the components, I guess there will be no problems if you checked the pinouts for your layout. However, if you copied the layout from someone else's, don't forget that pinouts for TO220 drivers are opposite to TO126 and of course pinouts for Japanese T092 style transistors are quite different to US and European (BC,BD etc). Google the pinouts if you are uncertain or have no data on your components.
If the VAS transistor or drivers are reversed from what they should be, either stage will be operating incorrectly or is now damaged. Check the voltage from base to emitter (Vbe) on all transistors and they should all be ~ 0.65V if working. Check also for shorts and other wrong connections.
You can be certain that the problems will be from errors in the PCB, errors in wiring or soldering or from using incorrect components. There's not much else that can be wrong.
well it is -28vdc measured with respect to 0V. I have double check the pinouts for the devices I used. I have the datasheet for all of them.
LTP are BC557A (facing the label it should be c-b-e)
VAS is BD139 (facing the label, E-C-B) and the drivers BD139/40
OP devices are 2SC5200 and 2SA1943 (B-C-E)
I will measure the Vbe of the trannies... and will update you soon.
I am using the circuit posted by Muff on post #70
Changed the miller cap to 68pf.
Thanks Ian.
LTP are BC557A (facing the label it should be c-b-e)
VAS is BD139 (facing the label, E-C-B) and the drivers BD139/40
OP devices are 2SC5200 and 2SA1943 (B-C-E)
I will measure the Vbe of the trannies... and will update you soon.
I am using the circuit posted by Muff on post #70
Changed the miller cap to 68pf.
Thanks Ian.
Last edited:
http://i1238.photobucket.com/albums/ff483/jmannile/DIGI125_schematic_small.jpg
I've linked Muff's copy of the schematic to use here for quicker reference.
For anyone interested, the power ratings are for 4R loads and R2 is 4k7, not 47k.
Note the comment that with 40V rails, a BC546 VAS transistor Q3, gets extremely hot and requires some form of heatsink. With 35V rails, it would still be very hot. It would be wise to revert to 100pF Cdom whilst troubleshooting as oscillation would bring its own heating problems.
I've linked Muff's copy of the schematic to use here for quicker reference.
For anyone interested, the power ratings are for 4R loads and R2 is 4k7, not 47k.
Note the comment that with 40V rails, a BC546 VAS transistor Q3, gets extremely hot and requires some form of heatsink. With 35V rails, it would still be very hot. It would be wise to revert to 100pF Cdom whilst troubleshooting as oscillation would bring its own heating problems.
Heh heh..My apologies for not reading the following post #62, which said:
Quote:
Originally Posted by Muff View Post
Just thought I would mention, in the schematic for the Digi 125, "R2" should be 4k7.
Errm, This resistor value for r2 quoted above, was a typo in a set of written instructions
from Greyham Dicker, on modifying the curcuit for 200 watts into 2 ohms.
The correct value is indeed 47k.
Quote:
Originally Posted by Muff View Post
Just thought I would mention, in the schematic for the Digi 125, "R2" should be 4k7.
Errm, This resistor value for r2 quoted above, was a typo in a set of written instructions
from Greyham Dicker, on modifying the curcuit for 200 watts into 2 ohms.
The correct value is indeed 47k.
OK, well the LTP will be imbalanced due to R3+Q3, but Q1 is doing nothing - something wrong there. By way of comment, BC557A are lowest gain options - perhaps standard or B,C suffix would be better in the LTP.
2N5401, BC556B, MPSA56 would also be OK there whilst debugging.
Q3, I would expect ~0.68V (don't ask me why, it just always seems to measure that with meters I use in amplifiers I work on!) This is running hot so yes, the voltage will drop below normal.
Also, check the Vbias across 3x1N4148 and again directly across the driver bases.
2N5401, BC556B, MPSA56 would also be OK there whilst debugging.
Q3, I would expect ~0.68V (don't ask me why, it just always seems to measure that with meters I use in amplifiers I work on!) This is running hot so yes, the voltage will drop below normal.
Also, check the Vbias across 3x1N4148 and again directly across the driver bases.
Last edited:
On a couple of occasions when I have had near rail voltage on the output, it turned out I forgotten to connect the ground correctly.
I only ever used 25V rails on my DIGI-125, even then, the VAS transistor got quite hot. I ended up using a TO-126 device like you.
Be careful if you play with the bias. I remember playing around with the number of diodes and the output got very hot at one stage. Use a big heatsink if playing around.
I only ever used 25V rails on my DIGI-125, even then, the VAS transistor got quite hot. I ended up using a TO-126 device like you.
Be careful if you play with the bias. I remember playing around with the number of diodes and the output got very hot at one stage. Use a big heatsink if playing around.
well, problem solve!!!
I replaced the BC557A with BC556B and the amp is now rockin' =)
I have made some modification also.
Made a bootstrap, added 1.2k (upper) in series with 2.2k (R6) then connected 47uF between them
Change C1 to 27uF and the result were very nice.. really better than chip amps!
I would call it Mid-Fi hahaha
Thanks Ian and Greg.
I replaced the BC557A with BC556B and the amp is now rockin' =)
I have made some modification also.
Made a bootstrap, added 1.2k (upper) in series with 2.2k (R6) then connected 47uF between them
Change C1 to 27uF and the result were very nice.. really better than chip amps!
I would call it Mid-Fi hahaha
Thanks Ian and Greg.
Last edited:
With R3=4k7 and R2=47K the voltage drop across R2 MUST be EXACTLY 20times the voltage drop across R3.
If not then the LTP is not in balance.
Any supply rail voltage above ~13V will unbalance the LTP. (and similarly supply significantly below ~13Vdc will unbalance the LTP).
Q1 will gradually turn off as the rail voltage is raised significantly above 13Vdc.
Ian F said
Sort the LTP BALANCE !!!!!!!!!
If not then the LTP is not in balance.
Any supply rail voltage above ~13V will unbalance the LTP. (and similarly supply significantly below ~13Vdc will unbalance the LTP).
Q1 will gradually turn off as the rail voltage is raised significantly above 13Vdc.
Ian F said
Sort the LTP BALANCE !!!!!!!!!
The detail of how any 2 identical amplifiers can be bridge connected may be seen diagramatically in this "Trimode" article":
Trimode Amplifier Bridging
Note that an opamp is used to invert the phase of the input to the second amplifier, which is necessary for bridging to work.
You also don't need a switchable arrangement, which is what trimode means.
There are some important points to understanding bridge-mode requirements first:
-Not all power amplifiers are capable or stable in bridge mode operation.
-Each amplifier "sees" only half the load impedance, so if you have a 4R speaker,
each amplifier will be driving 2R only. That is just too low for sensible use of Digi 125
and the amplifiers would need to be redesigned with a larger output stage to be useful at
low impedance.
-Dissipation and thermal tracking stability may not be well controlled here. You may
find you need to sacrifice amplifiers under working conditions to find out the safe limits
and this is expensive.
Personally, I don't agree with lowering load impedance to raise the power - this is the car audio solution for small equipment and simply raises the stress conditions on the amplifier, increases losses and reduces reliability. However, just about every guy new to audio wants to try it because bridging seems to deliver a free power increase from not just 2 but to 4 times and that looks wonderful until you realise how this power is delivered.
With the simple Digi 125, I would not try it unless you plan to use only with genuine 8R or higher impedance speaker systems (yes, many 8R speakers systems are actually 4-6R now because they are cheaper to make and play louder in a demo.) Note that all woofers dip well below their nominal rating in the lower midrange - around 3R is not unusual for 8R speakers).
Trimode Amplifier Bridging
Note that an opamp is used to invert the phase of the input to the second amplifier, which is necessary for bridging to work.
You also don't need a switchable arrangement, which is what trimode means.
There are some important points to understanding bridge-mode requirements first:
-Not all power amplifiers are capable or stable in bridge mode operation.
-Each amplifier "sees" only half the load impedance, so if you have a 4R speaker,
each amplifier will be driving 2R only. That is just too low for sensible use of Digi 125
and the amplifiers would need to be redesigned with a larger output stage to be useful at
low impedance.
-Dissipation and thermal tracking stability may not be well controlled here. You may
find you need to sacrifice amplifiers under working conditions to find out the safe limits
and this is expensive.
Personally, I don't agree with lowering load impedance to raise the power - this is the car audio solution for small equipment and simply raises the stress conditions on the amplifier, increases losses and reduces reliability. However, just about every guy new to audio wants to try it because bridging seems to deliver a free power increase from not just 2 but to 4 times and that looks wonderful until you realise how this power is delivered.
With the simple Digi 125, I would not try it unless you plan to use only with genuine 8R or higher impedance speaker systems (yes, many 8R speakers systems are actually 4-6R now because they are cheaper to make and play louder in a demo.) Note that all woofers dip well below their nominal rating in the lower midrange - around 3R is not unusual for 8R speakers).
The Digi125 is fine bridged into 8 ohms, mine has been in use now for 24 years, four modules bridged, driving your basic 3 way speakers with 15 inch bass drivers.
I have driven the amp very hard on many occasions without any problem,
and no component failures in the 24 years since I built it.
I spoke to Greyham Dicker over the phone once back then, for some advice and he recommended for best results, that I use Motorola transistors throughout the amp, which I did.
I have driven the amp very hard on many occasions without any problem,
and no component failures in the 24 years since I built it.
I spoke to Greyham Dicker over the phone once back then, for some advice and he recommended for best results, that I use Motorola transistors throughout the amp, which I did.
Rawbas's output transistors probably aren't in the old Motorola mould of tough, high current TO3 monsters. They could probably be bridged alright, but he would need to think of 2 pr to take advantage of the power increase, I'd think.
Amp1_zps100f041a.png Photo by rawbas | Photobucket
Amp1_zps100f041a.png Photo by rawbas | Photobucket
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.