Hi
This is not the best circuit in the world, IMHO. I doubt RF pick up is the issue from the input as there is a huge 3.9nF capacitor tying it down. Perhaps the trouble is in the back end because the BD711/BD712 drivers have the same frequency response as the output i.e. around 3-6 MHz each. This creates a significant phase shift above about 100kHz. There is a peak in the frequency response around 500kHz, hinting at instability. I would throw these drivers out and use a pair of BD139-BD140's myself, one pair for each output pair. Alternatively the high frequency drivers from ON semi (MJE15033/34) might be OK as a single pair replacement. The peak goes away with the BD139/Bd140 pair.
IF there is RF pickup then an output inductor/filter might help
John
This is not the best circuit in the world, IMHO. I doubt RF pick up is the issue from the input as there is a huge 3.9nF capacitor tying it down. Perhaps the trouble is in the back end because the BD711/BD712 drivers have the same frequency response as the output i.e. around 3-6 MHz each. This creates a significant phase shift above about 100kHz. There is a peak in the frequency response around 500kHz, hinting at instability. I would throw these drivers out and use a pair of BD139-BD140's myself, one pair for each output pair. Alternatively the high frequency drivers from ON semi (MJE15033/34) might be OK as a single pair replacement. The peak goes away with the BD139/Bd140 pair.
IF there is RF pickup then an output inductor/filter might help
John
Hi
30W into any speaker is 30W, so it does not matter whether this is provided by 2 or 8 ohms.
To drive 30W into 2 ohms you need about 7.8V rms and this will mean a current of 3.9A rms. To drive 30W into 8 ohms you need about 15V and 1.9A. So an amplifier will have to work harder to drive more current into a 2 ohm load than an 8 ohm load, as the load current is higher. The dissipation in the output transistors depends on the power supply voltage, and as the minimum voltage across the output transistors is usually around 5V (it can be lower if complementary pairs are used instead of Darlingtons) the dissipation is likely to be higher for the 2 ohm circuit even if operating at say +/-16V than the 8 ohm circuit (which needs to operate at about +/- 27V).
Which performs better just depends on how well designed each version is. You can design high current output stages (e.g. using triplets) but this adds to the minimum voltage needed!
John
30W into any speaker is 30W, so it does not matter whether this is provided by 2 or 8 ohms.
To drive 30W into 2 ohms you need about 7.8V rms and this will mean a current of 3.9A rms. To drive 30W into 8 ohms you need about 15V and 1.9A. So an amplifier will have to work harder to drive more current into a 2 ohm load than an 8 ohm load, as the load current is higher. The dissipation in the output transistors depends on the power supply voltage, and as the minimum voltage across the output transistors is usually around 5V (it can be lower if complementary pairs are used instead of Darlingtons) the dissipation is likely to be higher for the 2 ohm circuit even if operating at say +/-16V than the 8 ohm circuit (which needs to operate at about +/- 27V).
Which performs better just depends on how well designed each version is. You can design high current output stages (e.g. using triplets) but this adds to the minimum voltage needed!
John
That amp circuit is a decent oldfashioned circuit, which should work all-right, but never - never - never! with a supply of more than + / 30 Volt dc, as it is the maximum voltage, that MJ2955 can withstand. a R - C network on the output, of 10 Ohm in series with 0,1 mmFarad would be fine.
Regards Soeren Poulsen Denmark
Regards Soeren Poulsen Denmark
I used one of these amps in the 1980's for a disco.
It was very loud and served me well for many years.
I don't remember having problems with oscillation whether the input was connected or not although I think I put a pot on the front end so that might have calmed things down.
If I remember correctly it ran off +/- 45 volts.
I recently revisited the design for myself.
I changed the zener current sources to constant current sources.
I put in a current mirror on the LTP.
I changed all the transistors for modern transistors.
I had to increase the VAS capacitor a little to stop it oscillating with the newer wider bandwidth transistors.
Hi Soeren
The story about the 2N3055 is interesting. When RCA registered it the spec. for BVcbo was 100V. Typical BVceo was around 70. Now that it is made using epitaxial technology, the BVcbo and BVceo are broadly the same. In order to achieve 100V spec. this means that the BVceo is generally pretty high too. I have measured several samples of ON semi and ST 2N3055's and MJ2955's and mostly they are over 85V Bvceo if not over 100V.
I cannot guarantee that ALL devices will be this good (especially if counterfeits) but in general, I think +/-40V nominal is fine. I also suspect (but have no evidence) that most 2N3055's will be made on a line that also has transistors with 120, 140V or more BV's...
John
The story about the 2N3055 is interesting. When RCA registered it the spec. for BVcbo was 100V. Typical BVceo was around 70. Now that it is made using epitaxial technology, the BVcbo and BVceo are broadly the same. In order to achieve 100V spec. this means that the BVceo is generally pretty high too. I have measured several samples of ON semi and ST 2N3055's and MJ2955's and mostly they are over 85V Bvceo if not over 100V.
I cannot guarantee that ALL devices will be this good (especially if counterfeits) but in general, I think +/-40V nominal is fine. I also suspect (but have no evidence) that most 2N3055's will be made on a line that also has transistors with 120, 140V or more BV's...
John
Hi
If I recall correctly, the original used BD139-BD140 for the drivers. I do not recall when BD711/BD712 were specified, but no longer have any of the magazines that would have had the circuits. Using an HF driver with a medium frequency output gives the frequency response largely of the slower device, at least minimising the phase shift until the compensation capacitors have reduced the gain to below 1. With slow drivers and outputs a larger capacitor is necessary because the phase shifts are larger at the lower frequencies. In my earlier email I should have said that the kink in response I mentioned was in a simulation. I'd certainly replace the BD711 /BD712 drivers with something faster. These transistors were designed as outputs, not drivers.
John
If I recall correctly, the original used BD139-BD140 for the drivers. I do not recall when BD711/BD712 were specified, but no longer have any of the magazines that would have had the circuits. Using an HF driver with a medium frequency output gives the frequency response largely of the slower device, at least minimising the phase shift until the compensation capacitors have reduced the gain to below 1. With slow drivers and outputs a larger capacitor is necessary because the phase shifts are larger at the lower frequencies. In my earlier email I should have said that the kink in response I mentioned was in a simulation. I'd certainly replace the BD711 /BD712 drivers with something faster. These transistors were designed as outputs, not drivers.
John
Pair of Maplin 225WRMS amplifier modules
100 watt amp
An externally hosted image should be here but it was not working when we last tested it.
100 watt amp
An externally hosted image should be here but it was not working when we last tested it.
Last edited:
I just picked up a couple of the Maplin 150w bipolar amps.
Seller sent some info with it.
www.ckp-railways.talktalk.net/maplin225.htm
Fixed the first one, it had a dodgy plastic insulator on the BD711 driver transistor.
Took a while to find as it buzzed ok when turned off. Must have been arcing or shorting when turned on.
Seller sent some info with it.
www.ckp-railways.talktalk.net/maplin225.htm
Fixed the first one, it had a dodgy plastic insulator on the BD711 driver transistor.
Took a while to find as it buzzed ok when turned off. Must have been arcing or shorting when turned on.
Last edited:
Here is one I rescued and fixed off ebay.
Had a few components missing and a transistor plastic bush insulator had failed.
Had a few components missing and a transistor plastic bush insulator had failed.
An externally hosted image should be here but it was not working when we last tested it.
Here is the second one that I repaired.
I fitted new output transistors and they oscillated. SDOI increased VAS capacitor which calmed it down. I found that the new transistor had a much higher gain which was upsetting the circuit.
I fitted new output transistors and they oscillated. SDOI increased VAS capacitor which calmed it down. I found that the new transistor had a much higher gain which was upsetting the circuit.
An externally hosted image should be here but it was not working when we last tested it.
Hello everybody!
It is still so nice to see those fine old amplifiers. Regarding the Uceo of the output transistors, I believe, that newer transistors is selected sharper to the data specs, so, when we in the old days bought a MJ2955 it might had an Uceo mouch higher, than today. Running the amp at +/- 55 Volt dc, I would certanly look for a set of MJ15003/MJ15004, or the cheaper MJ15024/MJ15025, and as the driver something like a set of MJE15030/MJ15031.
Regarding the use of a set of BD712/713 as drivers, Dynacord uses a set of 2SA1302/2SC3281 in their Powermate 1000 mk1 and in their S900/S1200, which is very well sounding amps, and large!
It is still so nice to see those fine old amplifiers. Regarding the Uceo of the output transistors, I believe, that newer transistors is selected sharper to the data specs, so, when we in the old days bought a MJ2955 it might had an Uceo mouch higher, than today. Running the amp at +/- 55 Volt dc, I would certanly look for a set of MJ15003/MJ15004, or the cheaper MJ15024/MJ15025, and as the driver something like a set of MJE15030/MJ15031.
Regarding the use of a set of BD712/713 as drivers, Dynacord uses a set of 2SA1302/2SC3281 in their Powermate 1000 mk1 and in their S900/S1200, which is very well sounding amps, and large!
Last edited:
That was a weird pair of transistors! I have never seen that types! But they seems to be really fine things! What are the aprox prices for those devices? What did You use for drivers?
By the way: I wonder, if there is any advantage in making the output stage not in a darlington configoration, but like Rod Elliot like to do it, with the output transistors used as collector followers.
Regards Soeren Poulsen Denmark
By the way: I wonder, if there is any advantage in making the output stage not in a darlington configoration, but like Rod Elliot like to do it, with the output transistors used as collector followers.
Regards Soeren Poulsen Denmark
Last edited: