One More Tweedle Dee Build

[Blue Parrot]

Hi Folks,

I am currently getting near completing a Tweedle Dee amp build. I would like to thank Charlie Wilson and the other TAG members for taking the time to post photos and describe their builds and modifications.

Although numerous impressive Tweedle Dee builds have been highlighted here in the past, I wanted to post this build for the following reasons:

1) Planning to post this build forced me think a bit more critically about numerous build issues and potential problems while building the amp.

2) In the planning stages I struggled with decisions about certain components, in particular about the transformer set. I want to add to the existing data regarding the Magnetic Components 40-18021/40-18022 transformer set with either 5AR4CZ (GZ34 S) or 5Y3GT rectifier tubes.

I’ll start with the board. I prefer to use turrets over eyelets. Since I don’t have a drill press, I used smaller turrets which allowed using a hand punch to place holes in the G-10. Turrets were press fit and then flared on the back. Clearance can be fairly tight between some turrets and the input jacks. This is why the ground bus wire was run along the bottom of the board rather than on top. This approach, along with 0.25 in standoffs, gave plenty of room between the turrets and input jacks. Also, I used isolated input and output jacks and a segregated grounding scheme for this amp.

Selection of components was fairly straightforward except I was confused about the 0.0022 uF capacitor attached to pin 6 of V2. I used a MLR222K630 by NTE (from Radio Shack) although I wondered about this and came very close to using a Xicon 0.0022uF/630 metalized polypropylene capacitor instead. Are others using the MLR22K630 capacitor?

[Blue Parrot]

Here is the assembled amplifier. I considered using a TP40D power transformer from David Allen. This seemed to be a nice choice that allowed for a lot of options but I decided to stick with my original objective of building the Tweedle Dee, as presented by Charlie Wilson, and avoid getting sidetracked by designing in additional options. Therefore, I used the Magnetic Components 40-18021/40-18022 transformer set. I want to add more about the transformers in the next post.

[M Fowler]

Looks nice and tidy, good looking build.

I use eyelets for this type of chassis, not much room in there with all those input jacks.

Mark

[Blue Parrot]

Several builders have posted voltages obtained with the Magnetic Components 40-18021/40-18022 transformer set. Some obtained B+ voltages that agreed with Charlie’s report; however, it seems the majority of builds gave higher B+ compared to the desired 398 VDC. Below is a voltage chart using my line voltage (122.8 VAC) with a Sovtek 5Y3GT rectifier tube. The expected values in this table were for a 5AR4CZ (GZ34 S) rectifier and were compiled in an earlier post by M. Fowler. In general the voltages obtained were consistent with those reported by others. As shown in the table, I did obtain a B+ voltage that was slightly higher (417 VDC) than desired, even with the 5Y3GT. Using the limited data collected, I estimate that 118.5 VAC input voltage would result in the correct B+ (398 VDC) using the 5Y3GT rectifier tube with my AC line. This is a 3.5% drop from my actual line voltage.

I was curious how the 5AR4CZ (GZ34 S) rectifier would behave at different input voltages. The basic chart below is nothing earth shattering but does show that 115 VAC would result in the desired 398 VDC for B+ using this rectifier. This is a 6.4% drop from my line voltage. The heater voltage at 115 VAC is 6.12 VAC (rather than the intended 6.3 VAC). I think this heater voltage is probably high enough to use - could someone please verify that this voltage (6.12 VAC) is not too low for the heaters?

I plan to build a bucking transformer using a Hammond 166N6 transformer that is described in the link below. This will drop the line voltage either 3.7 or 7 %. This should give me the option of running either of the rectifier tubes and coming very close to the intended 398 VDC B+ voltage.

url=http://robrobinette.com/5e3_Modificatio ... ransformer

I did plug in the amp long enough to make sure it works with all of the 4 inputs. All the controls seem to work and the background noise level is very low. I verify, as others have noted, that this amp does get fairly hot. I will investigate this further once I get this installed in the cab. I am now tidying up some of the wiring and putting some finishing touches on the amp (knobs and tube retainers). I have ordered a speaker (Rocket 50’s are currently back ordered everywhere) and a cabinet (by John Mergili). I will post back when this build is completed – probably after burning in the components and getting used to the amplifier.

[Blue Parrot]

Thanks for the kind words Mark. Yes, it would have definitely helped having elf hands to assemble the parts and solder this in the limited chassis space!

[drew]

A NOS 5Y3 might bring the B+ down closer to what you're looking for.

[rogb]

Nice work, well done

[Structo]

Maybe the 40-18021 PT is made for 115-117vac?

[Blue Parrot]

Thanks Drew, sounds like something reasonable to try. I don’t know much about rectifier tubes, much less NOS rectifier tubes. Are there any specific NOS 5Y3s you might suggest that could fit the bill without breaking the bank? If so, I will certainly try.

Rogb – thanks for the kind comments. I have been following your current Tweedle Dee thread with great interest. BTW, love your cab!

Thanks Tom – I was thinking so initially too but the spec sheet for the Magnetic Components Classic Tone 40-18021 (included below) has a diagram that shows it has a 120 V primary with secondaries at 710 V CT, 6.3 V CT, and 5 V. Same sheet also states below the diagram in smaller print that the transformer gives a measured B+ of 396 V DC @ 100 mA load with a 5YGT rectifier. They don’t specify the primary voltage for the 396 VDC but I assumed they were talking about a 120 VAC input because of the diagram. Two things I can think of might explain the discrepancy between what the spec sheets state and observations by myself and others: 1) maybe they did use a lower VAC when measuring B+ and did not state this (although I would find this hard to believe), or 2) maybe they used a NOS 5Y3 that resulted in a lower B+ as Drew suggested above? It is confusing to me why myself and others are observing a B+ somewhat higher than stated in the spec sheet.

Scott

[martin manning]

JJ 5Y3's have a large voltage drop like the old ones, whereas the Russian ones tend to act more like a 5AR4/GZ34.

[68strat]

"Same sheet also states below the diagram in smaller print that the transformer gives a measured B+ of 396 V DC @ 100 mA load with a 5YGT rectifier."

396 V DC @ 100 mA load = playing the amp @ moderate to Loud volume to draw mA / pull down the B+ ~ 396 V DC.

More than 396 V DC @ 'less than 100 mA load' = when the amp is idling (not playing or playing lightly).

Just thinking outloud, thanks for all the tweedle post here, cool lil amp.

[Blue Parrot]

Martin – I will order a JJ 5Y3GT to substitute for the Sovtek tube to see if this drops B+ down to be more in line with the expected - thanks.

Hi 68strat - thanks for your interest and comments. The observed IR drop across the 250R cathode resister was 23.9 volts indicating a 95.6 mA (total for both output tubes) at idle. So current at idle was not too far off the 100 mA load stated in the spec sheet. Therefore, in this case I don’t think that the higher B+ was from low loading on the PT.

[67plexi]

I used the MOJO756EX with fantastic results 330-0-330 .075mA

I hope this helps.


Steve.

[martin manning]

0.075 mA? Really? I see that's what it says on the drawing, which also shows filament voltage at 3.15 V on each lead. With respect to what? There is no center tap! Sadly, this is typical of Mojo's documentation.

[jelle]

0.075 mA? Really?

You will void the warranty with a load higher than that.