Reverb Switching Alternatives

[Fresh_Start]

I would like to modify my Princeton Reverb derivative so that the grid of the reverb recovery stage doesn't go out to the footswitch & back.  IOW it needs to be quiet with the footswitch plugged in.

I know next to nothing about switching and even less about solid state devices.  Relays seem to be the most popular switching methodology here, but there just doesn't seem to be room to retro-fit one in this build.

Based on a conversation over on AX84 and a little bit of research, I came up with these two alternatives.  

The first uses the opto-LDR switch that Doug sells.  If I could use a half-wave rectifier with one 3.15 volt leg of the heater winding, that would be even cleaner.  Unfortunately, it didn't look like there'd be enough voltage.  The LDR needs about 2 volts and the 1N4001 eats another volt.  I know that the LED is a current driven device but thought I needed a resistor to drop voltage across and the value ended up being a fraction of an ohm.  Also, I don't know if the pi filter is overkill in terms of filtering the supply for the LED.

I can put the opto-LDR unit right across the 220K grid return resistor for the reverb recovery stage and add the rectification & filtering at the poweramp end of the chassis.  If I don't need the pi filter, I could even squeeze the whole thing between V2 and V3 using the heaters off of V3 as my source.

Doug posted the specs on the Vactrol unit HERE and there's a link to detailed specs at the end of that post.  BTW I was shooting for about 30ma for the LED on the theory that less resistance is better when the switch is closed.

J-FET switches seem to be popular among some folks, including Kevin O'Connor.  The circuit shown here is an attempt to patch together several pieces out of The Ultimate Tone.  Since the bias supply already has a half-wave rectifier for negative DC, I figured I could take DC voltage that was already dropped somewhat from the HV level and partially filtered.  Merlin mentioned something about a "gate resistor" - hopefully that's what the 1 meg shown is.

I can easily put the J-FET power supply on a terminal strip near the bias board.  I'm not crazy about the idea of dangling the J-FET over the turret board, but there may be a way to secure the lead with DC so it's not flopping around in space.

The biggest benefit to the J-FET is the cost.  Here's one that might do the job:
http://www.mouser.com/ProductDetail/Fairchild-Semiconductor/2N5638/?qs=ljbEvF4DwONAUxvcrCHfLA%3d%3d

So, do either of these two circuits work?  

The cost differential is HUGE given that the JFET is $0.14 and the opto-LDR is $9.00, but the opto-LDR approach appeals to me more... just because I understand it I guess.

I'd greatly appreciate all comments, corrections, recommendations, etc.

Thanks,

Chip

P.S. I asked the folks on the AX84 for their thoughts on this too, so please don't get aggravated with me if you see it over there as well.

[sluckey]

Either one will work. I'd prefer the JFET just because of cost and simplicity. (Be aware that JFETs come in two flavors just like bipolars, N-channel and P-channel. To switch off an n-channel device requires a negative gate-source voltage (VGS). Conversely, to switch off a p-channel device requires VGS positive.) An alkaline 9 volt battery would last like FOREVER when powering a simple JFET switch. All you really need is a battery, resistor, and FET.

But really, all this just to turn the reverb on/off? Leo's method works very well, if you use shielded cable to the reverb footswitch.

[Fresh_Start]

Granted a shielded cable to the footswitch seems like a good solution.  2-conductor shielded cable isn't cheap though.  I got starting thinking about this after a guy on the AX84 forum said he had a Deluxe Reverb that was dead quiet until he plugged in his footswitch.  Kind of fun to brainstorm on anyway.

I should have specified that my intention was an n-channel J-FET.

Ignoring the reverb application, I'd still like to understand the pros and cons of these two switching methods. KOC pretty much says "LDRs suck" but doesn't elaborate.  Other guys swear by them, and some are "rolling their own" LDR switches for a much lower cost alternative.  So what are the advantages & problems with each switching method?  (Please ignore reverb footswitch application)

On the J-FET approach, the 68K dropping resistor was a W.A.G. based on a formula I didn't really understand in The Ultimate Tone.  Is that in the right ballpark?

Will adjusting the bias affect the voltage supply for the J-FET?  I was thinking "no, the top (negative) end of that first filter cap is AC ground", but on second glance I'm not so sure...

Is the 10uf filter cap sufficient?  This is a pi filter with a 68K resistor, that's why I thought a lower value would be OK.  KOC shows a 470uf cap, but no pi filter.  Hoffman uses two 4,700uf caps in a pi filter for his relay power supply, but that's with a tiny resistor bridging the two caps.

Is the 100nf cap necessary?  Again, this drawing was cobbled together from pieces in TUT 1.  KOC says something about the 1 meg gate resistor and this cap providing a specific switching time.  Is that important?

Thanks for your thoughts.  If there are other references on switching you can recommend, please let me know.

Cheers,

Chip