silvertone 1484 reverb pan repair notes:

[soundmasterg]

Nuke, you could be right. I was going off a few old threads at the Hoffman forum about this. There was the one below, and then another I can't find right now that said something about the 150V caps being really cheap because they were used for something else. In any case, they used to be very common back in the day and aren't anymore.

Here is one theory from a Hoffman thread years ago about the use of the doubler circuits. One guys asked some questions which I put in bold and the responses on that forum.

Why aren't smoothing resistors needed in parallel with each of those 100uf/150v caps?

I've seen such a schematic once, but for a doubling circuit off one HV secondary. Because this design uses "stacked" HV secondaries, bleeder, or balancing, resistors might cause trouble. Also, the doubling circuit, by its inherent design, already distributes voltage between the series caps. So bleeder resistors may be superfluous. Most grand old designs do not use them.

If you applied this approach to this transformer with two 115 volt secondaries, would that give you an un-loaded B+ of 644? [(4 x 115) * 1.4]

Yes. And you don 't need more than one HV secondary for the voltage aspect. You can keep piling up doubling circuits off of one secondary, but at the expense of: a) decreased filtering, and b) increased current draw.

Would using the voltage doublers cut the current rating in half as well?

Yes. That's no doubt why Silvertone used 2 windings, to distribute the extra current draw among 2 windings.

Note that the caps are rated for only 150V. In the old days extra tranny windings were cheaper than hi voltage filter caps & diodes, which may not even have existed in the ratings required. Hence the existence of voltage doubling circuits back then, and their demise with modern hi voltage caps & diodes.

Greg

[soundmasterg]

Here's an interesting way to upgrade the reverb on a Silvertone 1484/1485 if anyone is interested.

https://www.youtube.com/watch?v=TIcW7sbA0hg

It does sound better than stock!

Greg

[solderhead]

greg, my hat is off to you for all of the work that you put into that amp. I can't imagine how many hours of work it took to design and then perform all of those upgrades while working in such a tight space with parts that were wrapped and coiled, then soldered. de-soldering must have been a nightmare.

when I imagined how much work it would take to do even minor changes on my twin twelve chassis I was motivated to put the amp on a shelf and work instead on projects that would give more immediate payback. I have to say that I've never seen a silvertone chassis that looks more heavily customized with PTP wiring than yours. It's amazing. have you ever thought about documenting it with a schematic, or do you prefer to just stick with descriptions? I'm sure that there are people out there who would truly appreciate a schematic.

.

my recollection is that the voltage doubler PT arrangement was chosen for cheapness in production, and nothing else. there were economies of scale that were at play back in the old days that must be hard to recognize now, as the price comparisons today don't really apply to the component prices of yesteryear. years ago it was cheaper to produce interstage transformers than capacitors, so most audio amps of the era were designed with transformer coupling. in the big scheme of things, capacitive coupling is a relatively new thing.

almost 30 years ago I had discussed this with an elder who was a design engineer at Motorola in the 40s-60s. he has since passed on but his comments to me mirrored your observations -- that this twin twelve voltage doubler PSU was primarily designed to be cheap to produce.

I don't know if either of you are aware of this, but Mojo had developed a twin twelve type of amp kit several years ago, which is now discontinued. they had built a modified circuit (no reverb, no trem, conventional cap input Pi-type PSU and B+ interruption standby) into a JTM-45 type chassis, using a JTM-45 PT instead of a voltage doubler, and a Heyboer reproduction of the funny little Silvertone OT. It was just channel 1 and channel 2, each with VBT controls. It was built using the basic JTM-45 chassis layout though it used turret board construction. At this point the custom Mojo chassis are long gone but the turret boards are still on their web site. I had always thought that the Mojo kit would be a good platform for experimenting with the circuit while avoiding the nightmare of working in a deep rat's nest of hard-wired PTP construction. somewhere around here I've got the old Mojo JTM-type chassis and OT just because I thought it'd be easier to realize the benefits of the unique qualities of the circuit while avoiding the hassle of working in the original chassis. like many things, that project is sitting with a lot of other spare parts that are on the back burner while more pressing projects have been commanding my attention.

[solderhead]

Here's an interesting way to upgrade the reverb on a Silvertone 1484/1485 if anyone is interested.

https://www.youtube.com/watch?v=TIcW7sbA0hg

It does sound better than stock!

Greg

Interesting video. when I was thinking about adding a fender-style reverb to a twin twelve, i had only thought about using a long-spring reverb tank. why? just because i like long-spring tanks so much more than short-spring tanks that I never would have considered a short-spring tank. with a long tank i thought i'd have to commit to mounting the tank above or below the chassis and that fitment would be a problem. it had never occurred to me that the short tank could fit iniside of the chassis if mounted sideways. and it sounds really good too.

nuke, i think that at least one reason that these amps sound so different from the garden variety FMV amps is that they use a bridged t-filter type of tonestack instead of the ubiquitous FMV tonestack. they really do sound different. in that respect they sound more like a baxandall or a james tonestack than an FMV, and the tone is reminiscent of those amps of the early 60s that used similar frequency shaping circuits. one person had commented in response to the YT video that the amp (with reverb) reminded them of the early Rolling Stones. IIRC they were using amps with similar tonestacks at the time.

[nuke]

I think we're a little too quick to jump on the "cheap" wagon. While the amp is obviously highly cost reduced, and it sold for literally 40% of the price of a similar Fender amp, all the engineering and design of the circuit is highly competent. Whoever designed the amp, absolutely knew what they were doing, in other words, a real, competent, electrical engineer did this work.

Sears was pretty much the Amazon of its day. They had a lot of power as a retailer and like today's Amazon, they probably had no mercy whatsoever with their vendors, both on price and on warranty issues. Customer got mad, complained enough, they got fixed or replaced, at the vendor's expense, in or out of warranty.

In realities of the day, spring reverb was protected by a bunch of patents and it was expensive to implement and probably required a license. The Danelectro reverb was almost certainly designed to get past the patents and they had implemented it that way for some time already. I think that really explains that one.

Given the cost containment and reliability concerns for warranty claims, the engineers designed very conservatively for reliability and to eliminate 'critical' tolerance parts. There are no 5% tolerance resistors, unlike Fender's designs. But the circuits were designed to work well over wide tolerances, and they do.

On the PSU section, some research on the parts and old catalog pricing, the BOM cost seems like a wash. The 100@150v and 20@500v electrolytics were about the same price in the early 1960's. The silicon rectifiers were the same ones that Fender used, 400PIV 0.75a. That works out to 140v RMS. Fender opted to string 3 in series (6 total) with a 640vac CT (320-CT-320) transformer along with a couple of 20uf@500v caps (or 600v). The 1484 (and the 1485) used the two stacked doubler supplies. That's 4 diodes and 4 caps. Weirdly enough, the caps and diodes cost each were about the same back then. It works out 6 diodes and 2 caps, or 4 diodes and 4 caps, which is 8 parts in the BOM of around the same price. I don't think the transformer cost was all that different either. Fender has filament + 3 windings on the secondary (every tap is a winding). The 1484 is also 3 windings, 2x for HV and 1x for bias. The HV windings were lower voltage, but probably higher current (thicker gauge wire).

fender-bandmaster-psu.jpg

silvertone-psu.jpg

One advantage for reliability in the Silvertone amps (1483, 1484 and the high power 1485) is running the screens at a reduced voltage. The 6L6GC was still kind of a new idea, the 5881 or 6L6WGB were the established high power tubes. The ratings on the 5881 and WGB all said to run them on lower g2 supplies.

The more I look at it, the more it seems me that they saw this as a big double-win: Get rid of the tube rectifier to save chassis space (big cost savings), get a nice, stiff, reduced voltage node for the screens supply and improved reliability in the output section.

I'll get into thoughts on tone in a follow up post. The tone stack is definitely part of it, but the gain structure is also a big factor.

[soundmasterg]

greg, my hat is off to you for all of the work that you put into that amp. I can't imagine how many hours of work it took to design and then perform all of those upgrades while working in such a tight space with parts that were wrapped and coiled, then soldered. de-soldering must have been a nightmare.

when I imagined how much work it would take to do even minor changes on my twin twelve chassis I was motivated to put the amp on a shelf and work instead on projects that would give more immediate payback. I have to say that I've never seen a silvertone chassis that looks more heavily customized with PTP wiring than yours. It's amazing. have you ever thought about documenting it with a schematic, or do you prefer to just stick with descriptions? I'm sure that there are people out there who would truly appreciate a schematic.

Solderhead, thanks for the props. I'm pretty sure everyone else who has considered modding one of these amps (or the 1484) quickly changes their mind, so yeah this one is likely the most modified around because I'm a glutton for self-induced punishment haha. I did write up a schematic and layout for myself, but it is on paper, and it is full size, so it is very large and not easily scannable. I haven't had the time to draw up what I did in a computer based program either. Maybe one of these days. I think from start to finish working on it here and there, all the mods on this and making the new head box and the 2x10 cabinet took me maybe 9 months or so? That was very early on when I was getting started working on tube amps, in fact it was the first amp I modified. It turned out well, but yes it was quite a lot of work. Planning in advance was extremely helpful. I don't enjoy working in these amps that much due to all the leads wrapped around all the terminals, and de-soldering isn't fun at all. The lack of space and the very poor layout also make it very hard to increase gain or mod things without hum popping up. I did everything I could to reduce hum in this amp, from lifting the heater center tap with a positive voltage, to using shielded wire to isolating and controlling grounds. It is still a little noisier than say a Fender Deluxe Reverb, but you can only do so much. In stock form they are actually pretty quiet, but they do have low gain in stock form. I did learn a lot from this project too so that was helpful to other things I've done since.

Greg