Hi all,
First post here and thank you in advance for the wealth of knowledge that I have been able to gain thus far.
I have a Cambridge A500 that no longer produces sound. The green Power LED comes on as well as the two red LEDs inside the unit. I expect that the problem here is (at least) the very obviously dead SAP15 P0 Darlington. I understand that sourcing a replacement set can be challenging and there may be downstream resistors that may also have been taken out.
So 2 questions if I may:
1) Is this worth repairing and if so, are these parts readily available?
2) This is just an assumption about the Sanken SAP but is there anything else I could check to verify this?
Thanks!
First post here and thank you in advance for the wealth of knowledge that I have been able to gain thus far.
I have a Cambridge A500 that no longer produces sound. The green Power LED comes on as well as the two red LEDs inside the unit. I expect that the problem here is (at least) the very obviously dead SAP15 P0 Darlington. I understand that sourcing a replacement set can be challenging and there may be downstream resistors that may also have been taken out.
So 2 questions if I may:
1) Is this worth repairing and if so, are these parts readily available?
2) This is just an assumption about the Sanken SAP but is there anything else I could check to verify this?
Thanks!
Oh dear. Absolutely and totally obsolete SAP**'s. Any from fleabay are probably fakes or unreliable. You may get lucky.
It is feasable to attempt to use standard Darlington complimentary pairs but good luck with the bias chain!
The SAP's don't like overload issues r shorted speaker leads.
It is feasable to attempt to use standard Darlington complimentary pairs but good luck with the bias chain!
The SAP's don't like overload issues r shorted speaker leads.
Thank you @JonSnell Electronic , you may have answered both of my questions in one! Shame as it was a cracking little amp. I've had enough challenges with bias on my valve guitar amplifiers so this may be a step too far.
The problem with replacing the transistors is it will likely just blow again.
What I do is remove outputs and feed output driver stage back into LTP.
This at least gives something that will turn on without blowing fuses.
But first a check of all components in the amp is required, especially VAS/bias stage.
A tedious job if its a big amp but it will find offending component(s).
What I do is remove outputs and feed output driver stage back into LTP.
This at least gives something that will turn on without blowing fuses.
But first a check of all components in the amp is required, especially VAS/bias stage.
A tedious job if its a big amp but it will find offending component(s).
Well, you can always sell it for spares, perhaps it will salvage another amp. 🙂
These proprietary output devices are a joke. They may have their advantages in terms of in-circuit tolerances, lower offset and parts count but when they become obsolete, theres no replacement, turning your amp into landfill.
At least you can sub a different output transistor for most discrete output devices, but these integrated transistor packs make it impossible to keep the amp going if it ever fails like the OPs amp did. I remember when Sony did the MN/MP2488 outputs on their receivers. These were expensive devices to replace but the receivers and amps they used these in were cheap.
Now with integrated class D being so cheap and improved, it makes entry level class AB amps obsolete, even though AB will sound better in most cases.
At least you can sub a different output transistor for most discrete output devices, but these integrated transistor packs make it impossible to keep the amp going if it ever fails like the OPs amp did. I remember when Sony did the MN/MP2488 outputs on their receivers. These were expensive devices to replace but the receivers and amps they used these in were cheap.
Now with integrated class D being so cheap and improved, it makes entry level class AB amps obsolete, even though AB will sound better in most cases.
Not sure whether this link might be of any use to you.
https://www.diyaudio.com/community/threads/alternatives-to-sanken-sap15-in-an-arcam-a65.168728/
https://www.diyaudio.com/community/threads/alternatives-to-sanken-sap15-in-an-arcam-a65.168728/
Would that be a technical appraisal of the design 😀
Tell you what I would do with this.... pretty much as @JonSnell Electronic mentioned and fit something like TIP141/142 Darlington's. Add something like 0.33 ohm emitter resistors. Bolt a TO126 NPN device (like a BD131) to the heatsink to use as a vbe multiplier and add a preset directly onto the BD131 for bias control. Dead easy to do, keep it as simple as possible. Do the same for both channels and then see where its at... fun project, nothing to lose and costs around £15 all in for all parts needed.
Parts count.
2 x TIP141
2 x TIP142
2 x BD131
2 x 1k preset
Insulating kits for all transistors.
Is there anything better for the Darlington's? The TIP141/142 are marginal on voltage but I suspect would be fine in practice.
Looks like the bias pot on the bad channel also released its magic smoke.
I also don't see any attempt in employing any sort of speaker DC protection. Maybe they would claim it negatively affects audio performance but i wouldn't trust the transformer secondary fuses to save your precious drivers.
I'd look for some mosfet based module and work them into the chassis, which doesn't look to be a complicated job in itself. You could do a pair of 250 watt Icepower modules, but the power supply for the tone controls is integrated into the older amp pcb. The juice isn't worth the squeeze IMO.
I also don't see any attempt in employing any sort of speaker DC protection. Maybe they would claim it negatively affects audio performance but i wouldn't trust the transformer secondary fuses to save your precious drivers.
I'd look for some mosfet based module and work them into the chassis, which doesn't look to be a complicated job in itself. You could do a pair of 250 watt Icepower modules, but the power supply for the tone controls is integrated into the older amp pcb. The juice isn't worth the squeeze IMO.
TIP141/142 are slow devices compared to Sanken s darlingtons that are typicaly 50MHz if not more,
beside they have huge Hfe at high current wich the TIP are desperatly lacking with the result of high distorsion.
Personaly i would repair it rather than installing new modules, and here the culprits are not the transistors
but the poor heatsink that dont even enclose all the transistors s dissipating area, the same wil hapen with
whatever other amp using those heatsinks as they are.
There s the 2SB1570/2SD2401 and a few others that would fit as repacement, altough you ll ave to add external diodes
stuck on the top of the transistor case.
I used those Sanken darlingtons extensively, for me they are among the best bipolar audio devices next to the legendary 2SA1302/2SC3281, they are very reliable but they are often used in exageratly high powered amps, FI the MP1620/MN2488 pair
is good for up to 30-35W but was used in 50W/channel amps.
beside they have huge Hfe at high current wich the TIP are desperatly lacking with the result of high distorsion.
Personaly i would repair it rather than installing new modules, and here the culprits are not the transistors
but the poor heatsink that dont even enclose all the transistors s dissipating area, the same wil hapen with
whatever other amp using those heatsinks as they are.
There s the 2SB1570/2SD2401 and a few others that would fit as repacement, altough you ll ave to add external diodes
stuck on the top of the transistor case.
I used those Sanken darlingtons extensively, for me they are among the best bipolar audio devices next to the legendary 2SA1302/2SC3281, they are very reliable but they are often used in exageratly high powered amps, FI the MP1620/MN2488 pair
is good for up to 30-35W but was used in 50W/channel amps.
wahab .... Is an hfe of 1000 not enough for you? I would say that is more than enough as the drive current available of at least 12mA that means 12Amps output drive!
Typical 4uS switch off and 0.9uS on, is pretty fast don't you agree.
Anyone who designs amplifiers knows that Tip142/147 are a good substitute.
Are you sure you are looking at the correct data sheet?
(Sorry Mooly, you show a pair of NPN with Tip141/142 80v and 100v.)
Typical 4uS switch off and 0.9uS on, is pretty fast don't you agree.
Anyone who designs amplifiers knows that Tip142/147 are a good substitute.
Are you sure you are looking at the correct data sheet?
(Sorry Mooly, you show a pair of NPN with Tip141/142 80v and 100v.)
Attachments
Thanks Jon, My mistake, I was going off memory. I think the TIP's would be for this. CPC have both these.
The TIPs should work, for the rest all depend of the amp design, if they are driven directly by the VAS, as is often the case,
then they present an input impedance of 8k at 5A output current on a 8R load, Sanken such devices have typically a Hfe of 12k@5A.
So the TIPs could yield quite more distorsion, and the fact that the VAS can deliver 12mA change nothing, as it is the output stage input impedance that most often define the VAS gain, and 12x higher Hfe mean 12x higher OS input impedance.
then they present an input impedance of 8k at 5A output current on a 8R load, Sanken such devices have typically a Hfe of 12k@5A.
So the TIPs could yield quite more distorsion, and the fact that the VAS can deliver 12mA change nothing, as it is the output stage input impedance that most often define the VAS gain, and 12x higher Hfe mean 12x higher OS input impedance.
@wahab I dont have the specs in front of me on the MN/MP darlingtons but I was under the impression they were 50w devices with 200 khz BW. The 3281/1302 were great Toshiba devices. I've used hundreds of these and was impressed how well they did, even without being derated on smaller heatsinks. I remember buying these things in lots of 100s and beta matching them. They were already very close among the same batches.
@profiguy, the MN2488/MP1620 are 150V/10A/150W with gain at 7A of 5-12k for the O lower grade, 6.5-12k for P grade
and 15-30k for the Y upper parts.
Ft are respectively 55/50MHz, set apart for the power/current/voltage they are comparable to the 2SA1302/2SC3281
and their usualy agregated drivers 2SA968/2SC2238 as far as we re talking audio perfs.
and 15-30k for the Y upper parts.
Ft are respectively 55/50MHz, set apart for the power/current/voltage they are comparable to the 2SA1302/2SC3281
and their usualy agregated drivers 2SA968/2SC2238 as far as we re talking audio perfs.
I was mistakenly referring to the max reliable power you could get from one complementary pair. Some will try to get 80w from a pair of 120w devices and even 100w from a pair of 150w TO3P. Rail voltage didn't matter so much using just one pair.
Sony used a pair of MN/MP to get over 80w/8 ohms and others commonly pulled 100w/8 ohms from a discrete pair of 2SA1302/ 2SC3281. That won't last long playing it loud for 30+ min without any external cooling, so no good for your typical teenager's house party. Forget about trying 4 ohm loads.
I've found that most TO247 were good for about 150w disappation and the smaller TO3P package could handle 120w. Many of the amps designed by the common audio brands ran on the edge of reliability using just one pair of outputs into anything less than 8 ohms. If you dared to run a 4 ohm load at medium output level, they'd quickly go up in smoke, even with the protection circuit doing its thing. You also didn't dare push the bias voltage across the common 0.22 R emitter resistors higher than a few mV, otherwise the amp would overheat quickly. Sometimes trying to reduce the amp's crossover distortion to a decent level required a bias setting which would make the amp overheat at modest power levels.
I understand the TIP142/147 aren't particularly the best output devices around, but some older car amps use them. I have a couple of Sony 2 ch amps with these outputs in my truck. They manage a solid 70w/ch into 4 ohms and 140w bridged 4 ohms. At first I didn't expect much for $50 each, but I was shocked when I heard them for the first time. They were just as good as any of the other common brands ie. Pioneer, Kenwood, Alpine, etc. While not quite as polished in the treble as my older Fosgate Power series, they're still clean and analytical - certainly better than many of the common brand class D amps I've tried.
Sony used a pair of MN/MP to get over 80w/8 ohms and others commonly pulled 100w/8 ohms from a discrete pair of 2SA1302/ 2SC3281. That won't last long playing it loud for 30+ min without any external cooling, so no good for your typical teenager's house party. Forget about trying 4 ohm loads.
I've found that most TO247 were good for about 150w disappation and the smaller TO3P package could handle 120w. Many of the amps designed by the common audio brands ran on the edge of reliability using just one pair of outputs into anything less than 8 ohms. If you dared to run a 4 ohm load at medium output level, they'd quickly go up in smoke, even with the protection circuit doing its thing. You also didn't dare push the bias voltage across the common 0.22 R emitter resistors higher than a few mV, otherwise the amp would overheat quickly. Sometimes trying to reduce the amp's crossover distortion to a decent level required a bias setting which would make the amp overheat at modest power levels.
I understand the TIP142/147 aren't particularly the best output devices around, but some older car amps use them. I have a couple of Sony 2 ch amps with these outputs in my truck. They manage a solid 70w/ch into 4 ohms and 140w bridged 4 ohms. At first I didn't expect much for $50 each, but I was shocked when I heard them for the first time. They were just as good as any of the other common brands ie. Pioneer, Kenwood, Alpine, etc. While not quite as polished in the treble as my older Fosgate Power series, they're still clean and analytical - certainly better than many of the common brand class D amps I've tried.
Unfortunately, all the Sanken darlingtons are going MOQ. The old school 3 MHz types do not always drop straight into a given amplifier. Where loop gain is high (>60 dB), they can cause an amp to oscillate if it was compensated for faster types. That’s because the integrated driver transistor is also slow. In more advanced circuits, the drivers (and predrivers if present) need to be high fT sustained beta types. The outputs can often be substituted with 3 MHz types, but not the drivers. In any case distortion does suffer. Whether one can hear 3MHz outputs vs 20 MHz the jury is out, provided the drivers are fast “enough”, especially with a triple.
In an old school vintage topology, with a diff pair resistively loaded, a single ended, non-cascoded, non-darlington VAS with a 100 pf Miller cap, directly driving complementary darlingtons any old pair will work and will likely be indistinguishable. As long as they are strong enough to drive the load without letting out the magic smoke.
In an old school vintage topology, with a diff pair resistively loaded, a single ended, non-cascoded, non-darlington VAS with a 100 pf Miller cap, directly driving complementary darlingtons any old pair will work and will likely be indistinguishable. As long as they are strong enough to drive the load without letting out the magic smoke.