slow AC-30 clone build.

[Mark]

I won't build this amp as fast as some of the builders here, but I'll take my time and do it to the best of my abilities.

I haven't thought how I want the preamp to differ from a stock AC-30 or not at all.

As far as the power stage goes, I'm liking some of Greg Fryer's ideas on protecting the AC-30 power stage and power supply from the hell it goes through. I like the idea of using 270 ohm to 330 ohm screen grid resistors.

I should point out that I will be using a 5AR4 rectifier instead of silicon diodes used in Greg Fryer's circuit.

I'm happy to hear thoughts and suggestions.

IMG_0988.JPG

IMG_0989.JPG

[Mark]

I just had a thought about plate resistors on the 5AR4 per Merlin Blencowe's site Valve Wizard. I'll look into this as I'm not worried about voltage drop as Mercury Magnetic are suppose to have higher secondary voltages, thus the additional resistance could drop voltage and protect the valve rectifier, the heat from the resistors is an issue though.

http://www.valvewizard.co.uk/bridge.html

[Colossal]

I just had a thought about plate resistors on the 5AR4 per Merlin Blencowe's site Valve Wizard. I'll look into this as I'm not worried about voltage drop as Mercury Magnetic are suppose to have higher secondary voltages, thus the additional resistance could drop voltage and protect the valve rectifier, the heat from the resistors is an issue though.

http://www.valvewizard.co.uk/bridge.html

Hi Mark,

I'm pulling up a chair for your build. I have read all of Greg Fryer's notes on his work on Brian May's amps and love AC30s. May's amps are stripped down to just half a gain stage right into the Vox PI and power amp. Dead simple and effective. Very interesting, given how those amps have to perform. Regarding a suitable Mercury PT, if you are planning to use one from their Vox catalog, the voltages are not inherently higher unless you get a true clone PT that is wound for 110VAC secondaries from days of olde and run it on modern mains. The MMP-V30H-M Haddon PT for example is 285-0-285 and I believe with a 120VAC primary. That is an unloaded voltage so with a GZ34/5AR4 rectifier and under load (we'll assume 50mA/tube at idle x 4 tubes is 200mA for an AC30 build), 285VAC * ~1.20 = 342VDC B+.

EDIT: Turns out that PT sags way more.

I have Merc 18W Radiospares PT in a Lite IIb derivative which is 301-0-301 unloaded with a 120VAC primary. Under load, this sags to 289VAC. Rectified on 122VAC with a Mullard EZ81, I'm getting 356VDC B+, 350VDC right on the plates and 315VDC on the screens (dropped across 2k2 10W and then 1k screen resistors). Cathode voltage is 12VDC across a shared 150R 10W. With no tubes, the system floats about 400VDC (301VAC * 1.35). I used 50R 25W resistors for good measure on the secondaries for a Rocket build once.

I love the AC30 sound, especially with a Les Paul and that's a direction I've been moving for a long time after experimenting with various preamps and EL84 power amps. I do tend to like a slightly warmer biased PI, more in the Marshally 18W vein. The 18W derivative I've settled on is very, very warm and detailed using Radiospares PT and OT, no choke.

Have you picked an OT and an operating point yet? Do you have thoughts on the more Marshally 10n/470k output section versus the Vox 100n/220k? I find the Marshall power amp plenty bassy. The AC30/4 power section is interesting as well, considering the different AC30-type engines. I heard an EF86 front end into an AC30/4 power section and it sounded like Billy Squire in a box. Killer, singing/wheezing grind!

[didit]

I should point out that I will be using a 5AR4 rectifier instead of silicon diodes used in Greg Fryer's circuit.

Why deviate from an overall upgrade? Unless there's something additional a vacuum rectifier gives you, the GZ34 is just ~15W of extra heat and probable failure.

Best .. Ian

[Roe]

after extensive testing I prefer:
ss rectifier and CLC filter (5h, 300ma, hammond)
- 470R resistor between mains and screens filter nodes
- 100R individual screen resistors
-32+32uf on the mains, 33uf on the screens, 10uf in the preamp, except for the cathode follower (22uf? - don't remember)
- seperate ef86, top boost and normal channels
- global brownface bias trem in the power amp
- Larmar master vol
I really prefer the 50R bias resistor on the el84s

[pompeiisneaks]

As side data, I've built the AC30/4 and love the normal EF86 channel and play it the most. I do use the trem channel every once in a while because it's probably the coolest I've heard ever... but I don't need that effect too often. It's also a bit quieter than the Normal channel that the EF86 gain pushes into the PI.

~Phil

[Mark]

Why deviate from an overall upgrade? Unless there's something additional a vacuum rectifier gives you, the GZ35 is just ~15W of extra heat and probable failure.

Thanks for your reply Ian. I'm not specifically after a Brian May amp. I'm shooting for the vintage sound without the vintage issues.

[M Fowler]

Sounds like a great project.

I bought a bunch of parts from various Ampgarage members years ago. Chassis, PT, OT from one guy which I believe are Weber kit stuff. The components I've collected over the years as well as buying from some TAG members.

I ordered a Weber AC30 cabinet so anxious to see how it looks because some of their cabs are not very well tolexed, amateur at best.

I too will be going at a snail pace on this build. It will be traditional AC30 circuit and I haven't even decided which schematic to follow yet as I'm above my head in local amp repairs, new builds, and experimental amps. Could use some assistants.

Mark

[Mark]

I have chased the EL-84 sound for a while. I built a 18 watt Marshall which I still use when I play with a band. I do recall the voltages were high in that amp, but it hasn't given me any troubles though I don't use it heavily.

I have built a Trainwreck Rocket and there were a lot of discoveries with that amp. I worked out the big advantage of having a 300VDC rail is the output stage is running under 100% dissipation. The downside is the small DC voltage on the first preamp stage. The 100 ohm versus 1K screen grid resistors were different too. I had the 100 ohm and 1K resistors on a switch and I cold switched between them. To my ear the 100 ohm resistors had more bass response.

As far as the AC-30 rail voltage goes I would prefer not to exceed 320VDC as that is still okay-ish for the output stage. I think 340VDC would see too much stress on the amp for little benefit, though I could be wrong.

As far as output transformers, I'm going to use a Mercury Magnetic Haddin transformers. I've heard the Haddin and Woden transformers sound very good.

I haven't made decisions yet about tweaking the PI though I do recall there was a mod where people replaced the 22K power supply resistor to the PI for extra headroom. I can't say I've heard of people still doing that.

[Colossal]

Could use some assistants.

Mark

Mark,

We should talk

[Colossal]

I have chased the EL-84 sound for a while. I built a 18 watt Marshall which I still use when I play with a band. I do recall the voltages were high in that amp, but it hasn't given me any troubles though I don't use it heavily.

Mark,

As with all things AC30, this is an interesting point of discussion. My 18W Lite runs at 349VDC on the plates, but just 315VDC on the screens. This works out to 12.5W per tube and the amp seems perfectly happy. There is a 2k2 resistor in place of a choke and then 1k on each screen, so there is a healthy drop on the screen voltage and keeps the dissipation in check. AC30s have a 470-500R choke and just 100R for screens, so if the plate voltage is high, so are the screens and this invariably leads to these 140% of max diss figures we see in literature, especially with amps struggling to live on today's mains voltages. I have not seen a definitive plate voltage value for an AC30, anything between 320 and 340VDC. I would expect they should be lower, in the 320s.

I have built a Trainwreck Rocket and there were a lot of discoveries with that amp. I worked out the big advantage of having a 300VDC rail is the output stage is running under 100% dissipation. The downside is the small DC voltage on the first preamp stage. The 100 ohm versus 1K screen grid resistors were different too. I had the 100 ohm and 1K resistors on a switch and I cold switched between them. To my ear the 100 ohm resistors had more bass response.

I followed a lot of that discovery when information was coming out about the Rocket. I also have done quit a bit of experimentation with screen resistors and do agree that 100R sounds great. I think a bit warmer and fuller to my ear in this style of amp. I'm always chasing articulation and clarity while retaining the warmth. My latest 18W build went back to the convention of 1k though, because of the inherently high plate voltage. To go back to 100R I would use a different PT and lower the B+ overall. Also worth discussing too is the contribution of the choke to the overall tone of the AC30 versus the Trainwreck or Marshall 18/36W convention of 1k 25W, 2k2, or 1k "choke" resistors respectively.

As far as the AC-30 rail voltage goes I would prefer not to exceed 320VDC as that is still okay-ish for the output stage. I think 340VDC would see too much stress on the amp for little benefit, though I could be wrong. As far as output transformers, I'm going to use a Mercury Magnetic Haddin transformers. I've heard the Haddin and Woden transformers sound very good.

Their Haddon is very, very good. What are your thoughts on using the modded Heyboer Dynaco A470? These were very popular when guys were building Rockets. Vox went through Parmeko, Woden, Haddon, and other suppliers.

]I haven't made decisions yet about tweaking the PI though I do recall there was a mod where people replaced the 22K power supply resistor to the PI for extra headroom. I can't say I've heard of people still doing that.

I haven't heard of increasing the PI plate voltage as a mod, but only reducing the bias resistor to address harsher, fizzier sounding cold-biased distortion that some experience. On an 18W, I tried the Vox PI 1M grid leaks and 0.01uF/220k AC30/4 power amp values but went back to the Marshall values in both the PI and poweramp. My goal was not preservation of the "Marshall" sound, just tuning to the sound I was looking for. I have found a number of comments where guys have cited preferring the Marshall PI and power amp, when bolted onto a stripped down AC30 preamp such as the Normal, Top Boost cascade, or EF86 versions.

For me, I really like the sound of the single 12AX7 Normal channel preamp with no tonestack, or an EF86 with simple 6-way tone and Cut control. Just fantastic sounds.

[M Fowler]

I found out really fast from a new local guitar shop how much he like the stock Rocket I built over the Dr Z Zwreck which I put in a 120K plate resistor instead of the normal 220k which breaks up more.

He immediately said I don't care for that Zwreck as much as that Rocket Combo. He had been using that Rocket combo in his band.

Now that plate resistor change was on my part to clean up the amp for country and doo-whap kind of bands. I am going to go back in and change that Zwreck to stock 220K.

Mark and Dave there are so many things one can do to tailor your AC30 amp to your likings.

Mark

[M Fowler]

Could use some assistants.

Mark

Mark,

We should talk

Yes we should talk Dave.

[M Fowler]

I used to use 300-0-300 TW transformers in my 30 watt EL84 builds but like Mike Zaite said that's a little too hot Mark!
I agreed and now run 290-0-290 in quad EL84's and 260-0-260 in my 18w builts. They are much happier. I normally run 100R screen resistors.

[Colossal]

I used to use 300-0-300 TW transformers in my 30 watt EL84 builds but like Mike Zaite said that's a little too hot Mark!
I agreed and now run 290-0-290 in quad EL84's and 260-0-260 in my 18w builts. They are much happier. I normally run 100R screen resistors.

Hi Mark,

On your quad EL84 builds using a 290VAC PT, is that the Heyboer 36W PT or equivalent? I'm guessing that 290VAC is a loaded voltage. Are you using a choke or just the 1k or 2k2 screen dropping resistor? Do you know you B+ offhand? If that's a loaded voltage and GZ34 rectified, I'd guess about 290 x 1.24 = 360V B+? I had 100R screens in my last 18W prototype build and it sounded great. The Marshall school is a higher 340-350VDC plate voltage but lower screens to compensate and Vox is 310-320VDC and screens just a little lower. Both amps are biased at 100%+. The Trinity forum Lightning guys also use a high plate voltage but much lower screens for their Matchless sound.