Was curious as to how "safe" it is to use the H11F1, H11F3 devices in a tube circuit for switching/muting purposes? I know several commercial devices use these: the Framus guitar amplifiers for example use the H11F3 exclusively for switching/shunting. The datasheet shows max signal for H11F3 as 60v peak-peak and 30 p-p for the H11F3. I know the voltages they are switching are up to 300v throughout! That's 10x the rated breakdown on these devices!!!!
I purchased some and emulated the circuit and it performs perfectly. Why is this? I noticed two things: these optoFETs never switch DC, and they always have a large series resistance (due to the high impedance nature of a tube circuit).
Also the device is only operated ON and OFF. never between. When ON there's little voltage difference btw in-out (about 20-40ohms) so no problem. When OFF there is a large voltage difference, but the device is 3Gigaohms as well as 100k ohms in series as well, so no breakdown?
Is this going to cause premature wear on these devices? Or is this a case of a situation where the devices are operating in a safe manner, just a very specific case the datasheet neglects to mention?
I have to assume the breakdown voltage of a semiconductor device has "some" relation to a "breakdown current -lol"...
I purchased some and emulated the circuit and it performs perfectly. Why is this? I noticed two things: these optoFETs never switch DC, and they always have a large series resistance (due to the high impedance nature of a tube circuit).
Also the device is only operated ON and OFF. never between. When ON there's little voltage difference btw in-out (about 20-40ohms) so no problem. When OFF there is a large voltage difference, but the device is 3Gigaohms as well as 100k ohms in series as well, so no breakdown?
Is this going to cause premature wear on these devices? Or is this a case of a situation where the devices are operating in a safe manner, just a very specific case the datasheet neglects to mention?
I have to assume the breakdown voltage of a semiconductor device has "some" relation to a "breakdown current -lol"...
Can you supply a schematic of the circuit in question? It's hard to tell what's going on from the description provided.
The opto-isolator is a bit of silicon that will tolerate some excess voltage but not 10x its rating.
The audio signals, OTOH, are relatively small signals and likely within the limits of the isolator - and the audio is what you want to mute, I presume.
As long as the optoisolator is protected from seeing the DC bias voltages, it could probably switch the audio fine. AC coupling, for instance, would allow use of the isolator.
The opto-isolator is a bit of silicon that will tolerate some excess voltage but not 10x its rating.
The audio signals, OTOH, are relatively small signals and likely within the limits of the isolator - and the audio is what you want to mute, I presume.
As long as the optoisolator is protected from seeing the DC bias voltages, it could probably switch the audio fine. AC coupling, for instance, would allow use of the isolator.
Here is my example circuit. Yes it is the audio I'm switching, and it is AC coupled. Using a ULN2803 as an LED Driver.
So, even though the audio can(does) swing btw ground and the +350v B+ supply, it shows no sign of breaking down in the optoFETs.
I've seen commercial products use these in the same way, but it's crazy that they survive the 10x the rated spec.
Is Breakdown being prevented by the 220k resistor between the optoFET and the supply? Does anyone think this will eventually break the device?
Maybe I should perform an endurance test to see if it ever breaks down...
So, even though the audio can(does) swing btw ground and the +350v B+ supply, it shows no sign of breaking down in the optoFETs.
I've seen commercial products use these in the same way, but it's crazy that they survive the 10x the rated spec.
Is Breakdown being prevented by the 220k resistor between the optoFET and the supply? Does anyone think this will eventually break the device?
Maybe I should perform an endurance test to see if it ever breaks down...
Very good. Thank you for posting your schematic.
This is a classic series/shunt switch. When the switch is "on", the series optoisolator is conducting and the shunt optoisolator is not conducting. It should pass the signal with little attenuation, assuming the output connects to a high-impedance load (such as a tube grid).
When the toggle switch is closed, the series optoisolator turns off, opening the connection, and the shunt optoisolator turns on, grounding the output.
The series/shunt should give very good isolation when "off" - it's a classic circuit used quite often when switching RF (where leakage capacitance through a solid-state switch requires the shunt switch on the output to achieve maximum attenuation in the "off" state).
It looks like you're safe, at least with the circuit as shown. The 22nF capacitor blocks the DC from the previous tube stage, so the optoisolator only sees the audio waveform which should normally be well below the 30V in a driver stage (ahead of the power amp).
If the next amplifier stage uses fixed grid bias (that is, if there's a fixed negative voltage on the grid), you might want to make sure that the bias voltage does not exceed the optoisolator spec. Normally, 30V would be a large bias voltage - but check to be sure.
If the next stage used cathode bias, the grid will return to ground through a large resistor and will not cause the optoisolator any trouble.
In fact, you probably need the grid resistor (to ground) to fully bias the shunt optoisolator "off". So it is a good thing
From what I can see, the circuit will do what you want it to do and you don't have to worry about the high voltage. It's on the previous tube, but not on the optoisolator.
I suggest that you be sure there is a grid resistor on the output - assuming you are driving a grid with this circuit. When the circuit is "off" you don't want the grid floating - the grid on the next tube stage should be biased properly regardless of whether the switch is "on" or "off".
Best luck with your project!
~ Sam
This is a classic series/shunt switch. When the switch is "on", the series optoisolator is conducting and the shunt optoisolator is not conducting. It should pass the signal with little attenuation, assuming the output connects to a high-impedance load (such as a tube grid).
When the toggle switch is closed, the series optoisolator turns off, opening the connection, and the shunt optoisolator turns on, grounding the output.
The series/shunt should give very good isolation when "off" - it's a classic circuit used quite often when switching RF (where leakage capacitance through a solid-state switch requires the shunt switch on the output to achieve maximum attenuation in the "off" state).
It looks like you're safe, at least with the circuit as shown. The 22nF capacitor blocks the DC from the previous tube stage, so the optoisolator only sees the audio waveform which should normally be well below the 30V in a driver stage (ahead of the power amp).
If the next amplifier stage uses fixed grid bias (that is, if there's a fixed negative voltage on the grid), you might want to make sure that the bias voltage does not exceed the optoisolator spec. Normally, 30V would be a large bias voltage - but check to be sure.
If the next stage used cathode bias, the grid will return to ground through a large resistor and will not cause the optoisolator any trouble.
In fact, you probably need the grid resistor (to ground) to fully bias the shunt optoisolator "off". So it is a good thing
From what I can see, the circuit will do what you want it to do and you don't have to worry about the high voltage. It's on the previous tube, but not on the optoisolator.
I suggest that you be sure there is a grid resistor on the output - assuming you are driving a grid with this circuit. When the circuit is "off" you don't want the grid floating - the grid on the next tube stage should be biased properly regardless of whether the switch is "on" or "off".
Best luck with your project!
~ Sam
Thanks for your insight, RF!
However I'm not sure I'm satisfied. I understand that the opto's never see DC, however the AC signal that these are switching often times is well above the 30/60v P-P voltage... I am able to drive that tube stage into clipping and optoisolators still switch ok.
Check this out:
http://www.framus.de/media/manuals/Amps/Dragon/Dragon_V5.pdf
Here is a commercial product that uses H11F3m's throughout in a manner very similarly to the example given above. If you look at the output of V3b (the cathode follower with the 100k on the cathode), you'll notice there's two H11f3's switching between two channel controls. This is a high gain amp, and that last stage is very much driven to clip to the rails, even on the gentlest of settings. Those optos are taking a lot of peak AC!
So my question remains, how are they handling this? Is this poor design and they are slowly failing? Are they ok because there is no DC, and why is AC ok and not DC? Are they ok because there's little potential for much current flow(hence the "breakdown current" idea)? Are they pk because the datasheet is uber conservative?
I don't get it! And thanks in advance...
However I'm not sure I'm satisfied. I understand that the opto's never see DC, however the AC signal that these are switching often times is well above the 30/60v P-P voltage... I am able to drive that tube stage into clipping and optoisolators still switch ok.
Check this out:
http://www.framus.de/media/manuals/Amps/Dragon/Dragon_V5.pdf
Here is a commercial product that uses H11F3m's throughout in a manner very similarly to the example given above. If you look at the output of V3b (the cathode follower with the 100k on the cathode), you'll notice there's two H11f3's switching between two channel controls. This is a high gain amp, and that last stage is very much driven to clip to the rails, even on the gentlest of settings. Those optos are taking a lot of peak AC!
So my question remains, how are they handling this? Is this poor design and they are slowly failing? Are they ok because there is no DC, and why is AC ok and not DC? Are they ok because there's little potential for much current flow(hence the "breakdown current" idea)? Are they pk because the datasheet is uber conservative?
I don't get it! And thanks in advance...
Hi,
If you have a concerned about the switching voltage of the H11F1m you can try the Avago ASSR-5211. It can switch 600volts at .2 amps. You can use it as a AC or DC switching depending on how the output pins are connected. Here it is the link for the specifications.
Data Sheet - ASSR-5211, High Current, 1 Form A, Solid State Relay (MOSFET), (600V/0.2A/16Ohm) (158 KB)
If you have a concerned about the switching voltage of the H11F1m you can try the Avago ASSR-5211. It can switch 600volts at .2 amps. You can use it as a AC or DC switching depending on how the output pins are connected. Here it is the link for the specifications.
Data Sheet - ASSR-5211, High Current, 1 Form A, Solid State Relay (MOSFET), (600V/0.2A/16Ohm) (158 KB)
Thanks tauro, I think you are on the right path, the datasheet says it can't do it, and there are plenty of other devices that clearly can. No use risking it!
I found the CEL PS7141E-1A-A device as well. 400v p-p ratings, high linearity and costs less than a H11F1. I'll order a few different devices and see if I can hear any difference.
I found the CEL PS7141E-1A-A device as well. 400v p-p ratings, high linearity and costs less than a H11F1. I'll order a few different devices and see if I can hear any difference.
I wouldn't use the series switch. Opto-couplers are not linear devices and the series resistance of the "switch" will vary with signal level. This will introduce distortion. In a guitar application this may not matter, or if you use it on the clean channel, then it will... The other thing is that you don't actually need the series switch. Just shunt the output to ground to mute it. The tube won't mind as you're not changing the DC voltage on the anode. You're just clamping the AC component.
Personally, I'd still use a small relay, though. The ULN2903 should be able to drive a small reed relay.
~Tom
Personally, I'd still use a small relay, though. The ULN2903 should be able to drive a small reed relay.
~Tom
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.