I'm restoring a Pioneer SX-950 and I noticed that the protection relay has no snubber diode across the relay coil. I have never seen this and I have to wonder if anyone has encountered a blown protection relay driver transistor in an SX-950.
It's possible that Pioneer was concerned about the snubber diode slowing down the relay too much and indeed they do, but a 24 Volt Zener back to back with the snubber does wonders for that. The relay coil is 24V.
I do run into some strange design decisions made by Pioneer and others, and sometimes I think that they were gambling with the design based on cost but some of the issues seem to be oversights and even a lack of knowledge at the time that we now take for granted. It's difficult to be cetain about these things.
The SX-950 also has a DC nulling system that I have seen fail and cause the protection relay to trip. The SX-880/890 seems to be the worst for this but the nulling circuit in that case needs to be recalibrated as the temperature sensing transistor they used ages.
This is my first deep dive into an SX-950 and I'm sure that there are other quirks. It should be interesting. In general it's a pretty magnificent piece of equipment and I look forward to giving it a listen when it's finished.
It's possible that Pioneer was concerned about the snubber diode slowing down the relay too much and indeed they do, but a 24 Volt Zener back to back with the snubber does wonders for that. The relay coil is 24V.
I do run into some strange design decisions made by Pioneer and others, and sometimes I think that they were gambling with the design based on cost but some of the issues seem to be oversights and even a lack of knowledge at the time that we now take for granted. It's difficult to be cetain about these things.
The SX-950 also has a DC nulling system that I have seen fail and cause the protection relay to trip. The SX-880/890 seems to be the worst for this but the nulling circuit in that case needs to be recalibrated as the temperature sensing transistor they used ages.
This is my first deep dive into an SX-950 and I'm sure that there are other quirks. It should be interesting. In general it's a pretty magnificent piece of equipment and I look forward to giving it a listen when it's finished.
Last edited:
Actually I would say that missing diode is quite common in devices from 70-ies (and even later on) - nothing I would consider unusual.
You can easily add it.
Photos of your progress would be interesting to see 🙂
Service manual can be downloaded from e.g. ElektroTanya if one has no HiFiEngine account.
And yes, there are quirks that can be solved in there - but I will not spoil your trip of discovering them...
Last edited:
It seems the relay driver, 2sc1318 is a very reliable device even if there is no snubber across the relay coil.
I have a sx-950 from new, it’s pretty much original, no protection ckt issues. The only thing I have changed are, the volume pot, output transistors and drivers because we blew the originals toshiba devices a few times until I changed the output to Motorola mj15003,15004. The only current issue is those silver plated switches oxidizing and causing intermittents, for now I just toggle the switches multiple times each time I use it. It’s on the list for servicing 🙂
I have a sx-950 from new, it’s pretty much original, no protection ckt issues. The only thing I have changed are, the volume pot, output transistors and drivers because we blew the originals toshiba devices a few times until I changed the output to Motorola mj15003,15004. The only current issue is those silver plated switches oxidizing and causing intermittents, for now I just toggle the switches multiple times each time I use it. It’s on the list for servicing 🙂
Thanks for the feedback on this rsavas!
I'ts interesting that they get away without the snubber and I would love to know if they were aware that this speeds up the relay significantly. The snubber holds back the highly inductive relay coil for discharging. Where this counts is when you loose an output transistor and the ouput slams to one rail and tries to light your speakers on fire. Douglas Self and one of his friends learned from a destructive test that this takes less than a second at =/-40 Volts and the SX-950 is nearly 50V. The SX-950 also has peak current detectors that open the relay faster than the DC offset circuit for such emergencies.
If I remember correctly the 2sc1318 driver transistor is rated at 50 Volts Vce and has every right to be damaged from a relay. I did do some simulations however and found that this varies wildly with the type of relay and other factors. Apparently the Pioneer guys worked it out and is of no concern.
The SX-950 I'm restoring at the moment is my first of this model and It may be the best in many ways. I'm not into that level of power but the construction and layout and so much more is really nice in these.
I'ts interesting that they get away without the snubber and I would love to know if they were aware that this speeds up the relay significantly. The snubber holds back the highly inductive relay coil for discharging. Where this counts is when you loose an output transistor and the ouput slams to one rail and tries to light your speakers on fire. Douglas Self and one of his friends learned from a destructive test that this takes less than a second at =/-40 Volts and the SX-950 is nearly 50V. The SX-950 also has peak current detectors that open the relay faster than the DC offset circuit for such emergencies.
If I remember correctly the 2sc1318 driver transistor is rated at 50 Volts Vce and has every right to be damaged from a relay. I did do some simulations however and found that this varies wildly with the type of relay and other factors. Apparently the Pioneer guys worked it out and is of no concern.
The SX-950 I'm restoring at the moment is my first of this model and It may be the best in many ways. I'm not into that level of power but the construction and layout and so much more is really nice in these.
Why do you think that the snubber has anything to do with relay (disengagement) speed?
Now I'm thinking that they knew about the speed issues and avoided the diode if it looked like the relay coil wasn't likely to make harmful spikes.
I am trying to better document my projects with photos and I have started to take some. It's also an interesting project already and I would love to share what I find.
This project will also involve a major cabinet restoration with complete veneering. I would be curious if anyone knows what kind of wood the solid end pieces are. One of my is badly damaged and I'm on the fence about what to replace them with.
I just noticed the comment on speed madis64. The speed issue is related to the reverse Voltage being shunted and this slows down the discharge of the relay coil enough to matter. Where it's most important is a catastrophic failure that may destroy the speakers and the most noticable effect in normal use is the muting of the speakers at power off. The problem however with keeping the power down clicks silent is that that the AC failure detection has to be on the fast side and that will trip the relay during brown outs or even heavy appliances sagging the Voltage where the houshold wiring isn't the best or sharing too many currents.
One of my designs from 2013 detects the AC in less than 1 cycle and I haven't been annoyed with false triggering of the protection relay. The power amp using this circuit has been used for my music performances since 2014 and even at venues with sketchy wiring didn't cause a problem. The snubber by the way gets around the speed issue by using a 4000 series diode back to back with a 24 Volt Zener. This makes them close to no snubber at all but still assures the transistor's safety.
One of my designs from 2013 detects the AC in less than 1 cycle and I haven't been annoyed with false triggering of the protection relay. The power amp using this circuit has been used for my music performances since 2014 and even at venues with sketchy wiring didn't cause a problem. The snubber by the way gets around the speed issue by using a 4000 series diode back to back with a 24 Volt Zener. This makes them close to no snubber at all but still assures the transistor's safety.
And anybody has had any reported/documented/described real life problems caused by that phenomenon?
Edit:
It is your device and you can do whatever you want with it but it has more serious problems inside than that magnetic field 🙂
I ran into two electrolytics in my SX-950 that do not agree with my KU service manual. What I don't know is how to identify the revision level of the boards and the manual.
One of the capacitors is 2200 µF and the maual shows 1000 and the other is 47 µF and the manual shows 220. Kind of strange that they go in both directions. This could be related to cost and supply etc. considering how much I see of that sort of thing in these oldies but goodies. I have also been involved with spontaneous engineering change notices in industry. It's easy to see that someone may have been putting out a fire.
One of the capacitors is 2200 µF and the maual shows 1000 and the other is 47 µF and the manual shows 220. Kind of strange that they go in both directions. This could be related to cost and supply etc. considering how much I see of that sort of thing in these oldies but goodies. I have also been involved with spontaneous engineering change notices in industry. It's easy to see that someone may have been putting out a fire.
Without component numbers (from schematic) it is difficult to understand what you are exactly writing about.
Both parts are on the power supply board - AWR-101 and 1000(manual) to 2200(board) is C12 and the 220(manual) and 47(board) is C14.
I was also speaking generally to see of anyone knows of different revisions of the boards or the entire assemble etc. that might be the issue here.
I was also speaking generally to see of anyone knows of different revisions of the boards or the entire assemble etc. that might be the issue here.
The pioneer documents are usually for the first production runs. Any follow on production changes usually never get documented. So the rule of thumb is to replace with what’s installed by the factory and not what was documented. As you say component changes can be due to design or cost savings.
I am not sure what awr-101revision I have, I’d have to look, but I do have a Canadian kc model, some times there are minor differences between ku and kc models.
I am not sure what awr-101revision I have, I’d have to look, but I do have a Canadian kc model, some times there are minor differences between ku and kc models.
Thank you, that' the sort of insight I was looking for rather than specifics. I am looking over the circuit and considering if I am making the best decision or Pioneer in the way back machine is. I encounter both scenarios for sure.
Ok, so these are the ones:
For me personally both options would work. C12 was probably increased to reduce noise, C14 was probbly reduced to save costs - but who knows exactly...