JCM 800 Switch 2 Volumes via MOSFET

[deke99]

I installed the Weber relay kit yesterday and the switching functionality works great. Both channels sound as they should and the transition from 1 to the other is perfect. The only problem is that I have an unacceptable amount of noise injected into the signal path when the coil is energized. I used shielded wire (grounded at only one end) for all signal runs to and from the board. The output return path goes right to the grid of the next gain stage and is actually a much shorter run that it was before. I have no tone/noise issues when the coil is off, only when it is energized.

Is there anything that I can do to minimize this? One thought that I had is that my shielded wire has 2 wires and I use 1 wire for the input lug send and the other wire for the output wiper send, so both wires are carrying signal at the same time within the shielded wire of the active pot (while the other pot's shielded wire has no signal). Would it be better to only have 1 of the 2 internal wires 'active' at a time? Are there any other measures that I can take to reduce the coil hum?

Thanks to all!!

[Rick Erickson]

I don't see this kit on Ted's page, so I can't relay comment on the kit itself. The first rule of thumb is: the energized state should be the clean channel, or lower gain channel. The power suply feed to the relay should be as clean as possible. I usually use a 78XX series voltage regulator to feed my relays. The coil should have a capacitor across it to filter the hum and switching spikes. The value of said cap will depend on the DC voltage and current requirements of the coil. I don't think your problem has anything to do with the way you are running the signals through the interconnect cables, although there is always the possibility - especially in a high gain amp to introduce ground loops by altering ground termination points.

RE

[deke99]

Hey Rick,

Thanks for the reply, here is a link to the kit details on Ted's page:

https://taweber.powweb.com/store/chansw.htm

There is currently no capacitor across the coil and I'm having trouble finding a data sheet for the relay. The relay is a TTI TR99-6VDC-SB-CD. So it would seem that it is a 6V supply. How can I figure out what value would I need to put across the coil?

[Edit] Ok, I found a datasheet on a russian site for the relay and the the coil is 6VDC and 75mA nominal current. Is there a formula that I can use to figure out what cap to put across the coil?

Thanks!
Derek

[Rick Erickson]

I'd start with 100uF/16-25V (or higher) and see how it responds. If there is a noticable switching lag then try a smaller one. If there is no reduction in hum then try a larger one. I don't have a formula for this - just trial and error. You may need to add an extra RC filter stage before the relay to get the hum down. The cap across the coil is really just to reduce switching transients.

RE

[Robert M. Martinelli]

[( e.g. if the supply voltage is 6 VDC you can' t switch signals with an amplitude greater than 6 Vpp ). ]

Last time I took signal measurements in a Marshall preamp, and mind you this was back in 1978, I found signal swings in excess of 50V RMS, or ~ 150V P-P. I don't know how you plan to bring the signal level voltage down enough to make it work with a 6V P-P switch, and then bump it back up enough to drive the amp properly. This is and always has been a major stumbling block in tube amps attempting to use solid state switching devices. A relay does not have a "clip threshold", I would go that route.

RE

Hi Rick,
I understand your point, but if we plan to switch the signal before a triode' s grid the system could work; 12AX7s are usually cathode biased with a low voltage, ( say -1 to -2 VDC on the grid ) this means you cannot apply signals in excess of 1 to 2 Vpp on the grid of a 12AX7 without clipping, and the 46xx series can surely handle 2 Vpp signals without problems.**

On the other hand, relays don' t have any of the issues commonly associated with solid-state switching, the only drawback being the audible "bump" they produce when switching.


** If I' m missing something please feel free to explain it to me, the reason why I love this forum is that I can share the little knowledge I have while improving myself by learning by others who have a better knowledge than I have.

Best regards

Bob

[Steve Conner]

Well, the grid-cathode diode of a tube only clips the positive peaks. The negative ones can go as far below ground as you like, and in a heavily overdriven vintage amp circuit, they do. The amp is probably sounding pretty crappy by this point, and Ray Ivers actually published a mod that uses a zener to clip the negative peaks too, claiming that it improves the tone. But the point is that a solid-state switch would clip the negative peaks at -6v or -9 or -15 or whatever, just like Ray's zener, and that would alter the tone. Again, probably for the better, but tube purists will throw up their hands in horror.

Speaking of which, I've not been able to get information on the Mojo MOSFET switching kit. My worry is that MOSFETs like the 2N7000 have an internal diode between drain and source. Unless Mojo used two MOSFETs back to back, the diode would clip any swings below ground even with the switch turned off. If you added a DC bias to avoid this, the result would be huge pops and bangs when the switch was used.

Bearing all this in mind, I'd also recommend using relays or LDRs to save hassle. They're more expensive than 2N7000s, but that doesn't matter unless you're in mass production. The difference in cost for a one-off build is only a few hours of burger flipping

[deke99]

Thanks for the info Steve! I did end up going the relay route. I tried using one of Ted Weber's kit boards but I couldn't eliminate the noise that the energized relay coil introduced. So instead I fabricated a little PCB (I've build a fair number of stomp boxes) to power the relay and mounted that close to the heater supply of V1. I then mounted the relay (OMRON 5VDC DPDT) on the board upside down (legs in the air, someone here referred to it as the 'dead bug' technique?) So the power board is close to the heaters and the relay is close to the pots being switched. With this arrangement I'm not getting any coil noise which is great! The power board and relay are based on the Marshall Jubilee implementation and I have found that it works very well.

I'm still working on my signal isolation since I'm switching 4 total pots, 2 on, 2 off at any given time, so I don't have complete signal isolation on both the input and output to each pot. Here is the power arrangement that I used, courtesy of SDM from the Metro amp forum:

[Tubis]

That single mosfet to ground will clip the signal. Not good for the tone at all.
Use the mosfets to drive opto isolators instead.
I dont know if you have any voltage supply for those but heres basicly how you install two vactrol optoisolators that will have 10M off and lower than 1k on. The switching is soft without any noise

[deke99]

Thanks for the input Tubis. I'm actually pretty happy with the relay implementation that I'm using right now. I just A/B'd the amp against a DSL 100 and this modded 2204 took the cake hands down.

I still think that I can do better though. I'm thinking about adding a second relay so that I can isolate the signal on both sides of the 4 pots. Does anyone see a potential problem using the same power supply that I included in my previous post to power a second relay? I did a quick test and I confirmed that the supply did switch both relays, but I don't want to wire it all up without knowing if there is a potential issue with that setup.

The relays that I'm using are listed as 200mW power consumption 40mA current.

Thanks to all for your continued help and support!

Derek