Hi all,
I thought about posting this in instrument section, but I think this is more or less a Class A amp circuit question.
The electra type guitar distortions use clipping diodes at the output. If you use 1n34a diodes, you get an output which is roughly equal to the clean signal you are feeding into the distortion circuit. However, if you use diodes (like 1n4148s) with a higher forward voltage drop, the distorted signal is much hotter. So if you want to match your maximum clean signal rms voltage or amplitude with the distorted one, you either have to add gain to the clean signal or reduce the output signal. This can easily be done with DPDT switches or 3PDT switches etc. But I was trying to go as simple as possible. Of course you can run different paths for clean and distorted etc, but I'm looking to use 1 transistor and doing something simple.
Here is what I did:
1) Instead of running the diodes to ground, I ran them to the positive rail
2) This requires another capacitor to remove the raise in DC voltage at the ouput
3) Then you can switch the collector resistor.
4) I varied the collector resistors to get a clean gain and drive control. (I also included an alternative Drive control that I think will work a little better and be less touchy with a linear pot, it is shown to the right of the circuit)
This requires one simple switch. I plan on running these into 2 different tonestacks and switching tonestacks at the ouput with a DPDT switch.
Anyhow, I have not seen this design before, specifically, running the diodes to positive rail instead of the ground, and using the collector resistors for gain control rather than the emitter resistor.
Is there anything wrong with this type of design or any anticipated problems?
I was thinking maybe there needs to be a series resistor say 1-5k in series with the distorted drive control, but ltspice simulation shows it works ok without it.
I thought about posting this in instrument section, but I think this is more or less a Class A amp circuit question.
The electra type guitar distortions use clipping diodes at the output. If you use 1n34a diodes, you get an output which is roughly equal to the clean signal you are feeding into the distortion circuit. However, if you use diodes (like 1n4148s) with a higher forward voltage drop, the distorted signal is much hotter. So if you want to match your maximum clean signal rms voltage or amplitude with the distorted one, you either have to add gain to the clean signal or reduce the output signal. This can easily be done with DPDT switches or 3PDT switches etc. But I was trying to go as simple as possible. Of course you can run different paths for clean and distorted etc, but I'm looking to use 1 transistor and doing something simple.
Here is what I did:
1) Instead of running the diodes to ground, I ran them to the positive rail
2) This requires another capacitor to remove the raise in DC voltage at the ouput
3) Then you can switch the collector resistor.
4) I varied the collector resistors to get a clean gain and drive control. (I also included an alternative Drive control that I think will work a little better and be less touchy with a linear pot, it is shown to the right of the circuit)
This requires one simple switch. I plan on running these into 2 different tonestacks and switching tonestacks at the ouput with a DPDT switch.
Anyhow, I have not seen this design before, specifically, running the diodes to positive rail instead of the ground, and using the collector resistors for gain control rather than the emitter resistor.
Is there anything wrong with this type of design or any anticipated problems?
I was thinking maybe there needs to be a series resistor say 1-5k in series with the distorted drive control, but ltspice simulation shows it works ok without it.
Attachments
Last edited:
Very noisy (thump) and probably power amplifier damage as the diodes are switched to positive rail, there will be a high voltage on the power amp input until the capacitor charges up. Not good at all!
Simple RC Snubber Circuit
Not sure if this would help, but at first glance this seems plausible. Found this circuit/figure on Wikipedia.
(Also, I tried changing the power source to a square wave to get a sort of turning on/off simulation and putting 1n4001 diodes at the ouput. Same arrangement as clipping diodes only to ground and 2 series diode pairs to leave room so clean signal doesn't get clipped. That would limit the height of the spike I would think, but really wouldn't get rid of the audible "plop" as 2 volt spike is still high)
Not sure if this would help, but at first glance this seems plausible. Found this circuit/figure on Wikipedia.
(Also, I tried changing the power source to a square wave to get a sort of turning on/off simulation and putting 1n4001 diodes at the ouput. Same arrangement as clipping diodes only to ground and 2 series diode pairs to leave room so clean signal doesn't get clipped. That would limit the height of the spike I would think, but really wouldn't get rid of the audible "plop" as 2 volt spike is still high)
Attachments
Please critique. I think this fix should work.
Hi there,
1) I put a 100k resistor across the switch so the diode capacitors should be fairly well charged under all conditions, at least within the 7-9 volt range. 100k was the minimum resistance needed keep interactions between the channels negligible. I found this experimentally.
2) I also put a simple RC charging circuit with a time constant of .47 seconds when the power switch is turned on so it'll take about a second or 2 to get close to full voltage. This also acts as a voltage divider and lowers the effect power supplied to the circuit to about 8.5 volts. But that shouldn't matter.
3) When turned off I'm guessing the charging capacitor will drain out through emitter to ground.
3) I put a square wave as a voltage source and didn't find any spikes, so I think this fixed the problem. (Not sure quite how to model the flip of the switch, but I'm thinking this shouldn't be a major issue)
Please give any comments if you see any problems. I'm trying to avoid blowing up any amps!
Hi there,
1) I put a 100k resistor across the switch so the diode capacitors should be fairly well charged under all conditions, at least within the 7-9 volt range. 100k was the minimum resistance needed keep interactions between the channels negligible. I found this experimentally.
2) I also put a simple RC charging circuit with a time constant of .47 seconds when the power switch is turned on so it'll take about a second or 2 to get close to full voltage. This also acts as a voltage divider and lowers the effect power supplied to the circuit to about 8.5 volts. But that shouldn't matter.
3) When turned off I'm guessing the charging capacitor will drain out through emitter to ground.
3) I put a square wave as a voltage source and didn't find any spikes, so I think this fixed the problem. (Not sure quite how to model the flip of the switch, but I'm thinking this shouldn't be a major issue)
Please give any comments if you see any problems. I'm trying to avoid blowing up any amps!
Attachments
dual channel single 2n3904 guitar preamp
I think biasing the output and putting a cap from diodes to ground will take care of the transient high voltage output as below.
I think biasing the output and putting a cap from diodes to ground will take care of the transient high voltage output as below.
Attachments
Last edited:
Simple Channel Switching Schematic for Electra Type Distortion
Here is a different way to deal with these problems, I have simulated pulses built the circuit it seems to work fine. I added these IR diodes which clip around 1v, just as pulse limiters for the clean channel, the distorted channel does not clip it acts almost identically to the original electra design. It sounds much better with 1n34a's in place of 1n4148s but of course quieter. The 22uf cap also allows the diodes to be mixed and matched without adding any dc offset according to simulations.
This attached schematic seems to be the most straight forward way to flip an electra from clean to distorted and still take advantage of the gain. You could change the 2k pot to a 5k pot and put a 5k trimmer in parallel to adjust the maximum output level of the clean channel to match the distorted channel if you are using a preamp. Also I ran simulations and made a chart to help figure outputs with various diode pairs I will post separately.
Here is a different way to deal with these problems, I have simulated pulses built the circuit it seems to work fine. I added these IR diodes which clip around 1v, just as pulse limiters for the clean channel, the distorted channel does not clip it acts almost identically to the original electra design. It sounds much better with 1n34a's in place of 1n4148s but of course quieter. The 22uf cap also allows the diodes to be mixed and matched without adding any dc offset according to simulations.
This attached schematic seems to be the most straight forward way to flip an electra from clean to distorted and still take advantage of the gain. You could change the 2k pot to a 5k pot and put a 5k trimmer in parallel to adjust the maximum output level of the clean channel to match the distorted channel if you are using a preamp. Also I ran simulations and made a chart to help figure outputs with various diode pairs I will post separately.
Attachments
- Status
- Not open for further replies.