Hello. I am looking for ideas. I want to have the ability to check my tube amp DHT filament voltages with a built in voltmeter circuit. Connecting anything to the DHT filament can affect the sound. So my idea is to use a relay as shown in the attached circuit. Any ideas would be much appreciated.
Thanks,
Adrien.
Thanks,
Adrien.
Attachments
most Weston gauges are having big enough input impedance for not having affect on sound
and , DVMs are mostly up from 1M input impedance
whatever , if you insist on just temporary connection , I see no reason to use anything more complex then ordinary switch
and , DVMs are mostly up from 1M input impedance
whatever , if you insist on just temporary connection , I see no reason to use anything more complex then ordinary switch
switching the relay on and off itself would generate a more audible noise (a electromagnetic and a inductive one)
just put test jacks on the chassis to hook a dvm up to it when you want to check. Unplug it when done. I just use banana jacks. There cheap and it works well. And no sound interference.
Hi!
Put resistors into the leads between volt meter and filament terminals. Say 1kOhms.
If your voltmeter has an input impedance of 1M this won't affect accuracy. 1k will decouple the filaments which have lower AC impedance to ground. If your voltmeter has higher input impedance, you can use higher value series resistors.
Best regards
Thomas
Put resistors into the leads between volt meter and filament terminals. Say 1kOhms.
If your voltmeter has an input impedance of 1M this won't affect accuracy. 1k will decouple the filaments which have lower AC impedance to ground. If your voltmeter has higher input impedance, you can use higher value series resistors.
Best regards
Thomas
Sampling action of DVM actually imposes measurable noise on many power supplies, so I would certainly take care with those.
With any other solution, please take care to avoid long wires, as these may collect radiated noise.
Short wires to a test socket or switch should be trouble-free.
With any other solution, please take care to avoid long wires, as these may collect radiated noise.
Short wires to a test socket or switch should be trouble-free.
Thanks for all the replies. I put together the attached circuit diagram. It's a bit difficult to illustrate. I hope it makes sense.
The idea is to keep wires to the meters disconnected except when monitoring. Also to use analog so there is no clock noise influence. Monitoring the GM70 bias will be the default position. I will be using a 5 pole 11 position rotary switch. I found a really nice switch made in USSR.
The connections to the filaments will be short to the relay contacts. The relay normally closed position will be used to ground the long wires when not in use. The capacitance of the relay contacts is low so there should be minimal influence from this:
Capacitance (10 MHz)
1.5 pF (between open contacts)
1.0 pF (adjacent contacts)
1.7 pF (between coil and contacts)
I added some test banana plugs to hookup a DVM for accurate setting of the bias and filament voltages. I figured out an auto-ranging cct so that I could do this.
I believe this satisfies the requirement:
Monitoring filament voltage with minimal effect.
Adrien.
The idea is to keep wires to the meters disconnected except when monitoring. Also to use analog so there is no clock noise influence. Monitoring the GM70 bias will be the default position. I will be using a 5 pole 11 position rotary switch. I found a really nice switch made in USSR.
The connections to the filaments will be short to the relay contacts. The relay normally closed position will be used to ground the long wires when not in use. The capacitance of the relay contacts is low so there should be minimal influence from this:
Capacitance (10 MHz)
1.5 pF (between open contacts)
1.0 pF (adjacent contacts)
1.7 pF (between coil and contacts)
I added some test banana plugs to hookup a DVM for accurate setting of the bias and filament voltages. I figured out an auto-ranging cct so that I could do this.
I believe this satisfies the requirement:
Monitoring filament voltage with minimal effect.
Adrien.
Attachments
Your plan will only identify an out-of-tolerance voltage by manually doing the switching. Maybe you enjoy the thought of checking all your voltage levels when you turn the amp on - but in effect you are only checking for far less than 0.1% of the time the amp is on, and hence missing events that occur at other times.
Simplistically, you could add a self-powered LP358 type tolerance detector with opto-coupled output for many of those voltage measurements. Then you could achieve something like 100% checking opportunity, and even realtime power down or other form of disabling if a fault occurred.
Simplistically, you could add a self-powered LP358 type tolerance detector with opto-coupled output for many of those voltage measurements. Then you could achieve something like 100% checking opportunity, and even realtime power down or other form of disabling if a fault occurred.
You are correct, my solution has it's limitations. I do however monitor the bias current of the GM70 real time and have a protection circuit for that as well. The rest of the monitoring requirement was to check the voltages manually when the need arises. Like when the tubes are replaced for example. My amp was built utilizing any parts I had in my parts bin. As a result, it has a hand built vintage look and feel. This manual method of measuring the voltages using vintage analog meters and rotary switch fits right in with the rest of the amp.
After I posted the circuits yesterday, I realized I didn't need the LM339 comparator circuit. The auto ranging can be controlled from the rotary switch 12V pole. I will review the new circuit for errors before posting.
Adrien.
- Status
- Not open for further replies.