Akai 707 Preamp Modification

[untune]

Brilliant, all slotting into place now. I reckon a last tweak of operating points and values tonight and I should be ready to put it together and see how (if? ) it all works as expected.

Thanks for the link, I'll have a proper read of the article later. I suppose logically it doesn't matter if you're dropping more than a couple of volts per filter, though lower value resistors might have too high a corner frequency in conjunction with the cap... I've seen an example that dropped something like 9v per tube, might be better to go for something like that

Cheers
Lee

 

[PRR]

> 5K will probably be OK.

Agree. I may have been designing for lesser tubes where we struggle for +23dBm out. The EL84 can do 5 Watts (+37dBm!) as pentode, hardly a wee tube. And untune won't have London calling asking for more level on the radio network line. So even with the lesser triode power and much lower efficiency in Resistance Loaded, it should be a fine output.

 

[PRR]

> matter if you're dropping more than a couple of volts per filter

Scroggie 1949 is maybe more optimisation than practical. He assumes a 12uFd may be divided into four 3uFd and shows better performance. This might even have been true in the day when 8uFd was a big cap and 12uFd might well have been a heap of 3uFd. Today they charge you per can and 3u/6u/12u may all be the same price, making the 4-stage solution more expensive.

His point is valid forever: given this type filter, more stages is better than more uFd per stage. Looking at many factors, taking 30:1 reduction in each stage is a good starting point. You can't get 1000:1 in one stage because of wire and cap resistance. Less than 10:1 per stage is just a lot of stages. Expect to be 4 stages down by the time you hit your first (lowest level) stage.

Radiotron 3rd has a section on B+ chain filtering, very practical, but radio-conservative. He advocates for a 30% to 40% drop in the filter! This leads to large R (same cost) and small C (cheaper). However in radio we often know just how big the signal is (even have AVC to enforce that). Live musicians are not so predictable, we like lots of B+ available. Also where a 8uFd cap used to be a big costly box, 100uFd is now a cheap can.

I think you would do well to COPY the chain AKAI used, maybe doubling the caps in light of current prices. They were not fools.

Another happy guide for much modern tubewerk is 10% per filter stage. A 10% drop in B+ is hardly anything (if 270V is not enough, 300V won't be much more).

 

[untune]

Thankyou both for your input - I plotted everything based on ~250V/5K anode load and ended up with some values I was happy with. 120R on the cathode, 25uF cathode bypass (for now) and I've got a VTB2291 on the way.

You make a very valid point in regards to copying the original. As it happens, I recalculated with a drop of ~20v between stages - the new dropping resistor values are actually not a million miles from the originals. As suggested by Ian I upped the caps from 20uF to 47uF, so I'm feeling confident I'll have a good starting point at the very least.

The last thing I need to decide on is a mic input transformer! Preferred mics tend to sit at 150-200ohms but a few in my collection are 300 and 600 - is there an ideal option going into a triode 6AU6?

Cheers
Lee

 

[ruffrecords]

The last thing I need to decide on is a mic input transformer! Preferred mics tend to sit at 150-200ohms but a few in my collection are 300 and 600 - is there an ideal option going into a triode 6AU6?

Cheers
Lee

Tubes do not have great noise figures so they need all the help they can get. One way is to get some fairly noiseless gain up front with a step up transformer. 1:10 is surprisingly common. Cinemag, Jensen and Sowter make suitable types. I would not recommend skimping on the cost of this transformer. If you want to buy in the UK the Sowter 9610 would be a good choice.

Cheers

Ian

 

[untune]

I'd quite like to try the Jensen and Cinemag trannies, they seem popular with these types of project (the JT-115K-E especially) but hard to get over here and I have a thing for Sowters. Not cheap but worth it! The 9610 looks perfect so I've got one on the way - thanks for the suggestion Tons of room on the input side to mount it.

Looking forward to finally getting stuck into this - I'm going to try a couple of 100ohm resistors for an artificial centre tap on the heaters. Later models used a 300 ohm pot to balance it, I might give that a go too.

I've reduced down to two coupling caps - one between the first and second stage and one between second and third.  I'm unsure if I need one between third stage and OT, or if the is no DC as a function of the way the EL84 is arranged?

I reasoned that the high pass could be made by switching the tone pot into the circuit following stage one, with the coupling cap and subsequent earthed variable resistor creating the filter; opening the switch breaks the connection to earth/the pot.

I might put an attenuator on the output - or I might build a little H-pad in an external box and use that if I need it (http://www.preservationsound.com/?p=7095)

The VU meter is on hold for the moment - I have a nice meter I could potentially use, but it's DC and might need some work to get it in. I've also now got an unused B7G valve base from the 6AR5, and something like 0.44A of headroom on the heater tap, so I could potentially have another valve in there for some other purpose. I wondered if there might be a simple compressor or limiter possibility but it might be a bit beyond reach at this stage!

Cheers
Lee

 

[ruffrecords]

I'd quite like to try the Jensen and Cinemag trannies, they seem popular with these types of project (the JT-115K-E especially) but hard to get over here and I have a thing for Sowters. Not cheap but worth it! The 9610 looks perfect so I've got one on the way - thanks for the suggestion Tons of room on the input side to mount it.

I like Cinemags myself. By the time you pay shipping and import duty they are no cheaper than Sowters

Looking forward to finally getting stuck into this - I'm going to try a couple of 100ohm resistors for an artificial centre tap on the heaters. Later models used a 300 ohm pot to balance it, I might give that a go too.

Definitely worth a try. When the tape recorder was made, a signal to noise ration better than 40dB would have been good so there was less incentive to minimise heater or HT supply hum. For mic pres we expect much better performance.

I've reduced down to two coupling caps - one between the first and second stage and one between second and third.  I'm unsure if I need one between third stage and OT, or if the is no DC as a function of the way the EL84 is arranged?

In a class A circuit, the average current is independent of the signal level (unlike a class B where the bigger the signal the bigger the average current draw. However the current does still vary with the signal so it is important to think about where these short term current are going to be supplied from. A local reservoir capacitor helps supply these and reduce the demoand on the rest of the power supply.

I reasoned that the high pass could be made by switching the tone pot into the circuit following stage one, with the coupling cap and subsequent earthed variable resistor creating the filter; opening the switch breaks the connection to earth/the pot.

Sounds good to me.

The VU meter is on hold for the moment - I have a nice meter I could potentially use, but it's DC and might need some work to get it in. I've also now got an unused B7G valve base from the 6AR5, and something like 0.44A of headroom on the heater tap, so I could potentially have another valve in there for some other purpose. I wondered if there might be a simple compressor or limiter possibility but it might be a bit beyond reach at this stage!

You could add a 6922 (E88CC) configured as an SRPP and use it to drive a Vactrol  to include a compression stage. I did something similar in the poor man's tube compressor.

Cheers

Ian

 

[untune]

Good point actually. I rarely every buy anything from the States for the fear of getting stung on the duty but I suppose if it doesn't cost any more then it makes sense. Especially since Sowter have given me a lead time of 2-3 weeks!

The wiring is atrocious so my first task after dismantling and cleaning is to properly twist and map the heater wires around the chassis. This may be a daft question but is it possible to buy the shrouds/shields that you get on transformers or are they inherent to the design?

Understood regarding the filtering - I think in order to properly supply each stage with the correct HT I need to shift the dropping resistors and caps around from the original. So I'll have a 33K+47uF for stage 1 and a 5K6+47uF for stage 2, but since the HT doesn't need to drop for the third stage and comes from the first smoothing cap after the choke, will I need another smoothing cap or is it surplus?

I have seen your poor man's tube compressor and it looks like the perfect option! I'd need to swap the b7g for a b9g base which could be awkward as they're riveted to the chassis. I'm sure my dad's drill press could take care of that - in fact I was thinking of drilling them out and remounting with screws and rubber spacers for the 6au6s at the very least. I also saw a similarly modified akai/roberts machine that used the spare b7g for another valve.  I forget what it was (and its purpose) but I believe it was one half of something like a 12at7 in a single tube.

Cheers
Lee

 

[ruffrecords]

Good point actually. I rarely every buy anything from the States for the fear of getting stung on the duty but I suppose if it doesn't cost any more then it makes sense. Especially since Sowter have given me a lead time of 2-3 weeks!

That is pretty good. I just got some from CInemag which took about twice that.

The wiring is atrocious so my first task after dismantling and cleaning is to properly twist and map the heater wires around the chassis. This may be a daft question but is it possible to buy the shrouds/shields that you get on transformers or are they inherent to the design?

Are you talking mains transformers here  or audio ones?

Understood regarding the filtering - I think in order to properly supply each stage with the correct HT I need to shift the dropping resistors and caps around from the original. So I'll have a 33K+47uF for stage 1 and a 5K6+47uF for stage 2, but since the HT doesn't need to drop for the third stage and comes from the first smoothing cap after the choke, will I need another smoothing cap or is it surplus?

I think we already increased the one after the choke so that will probably be enough.

I have seen your poor man's tube compressor and it looks like the perfect option! I'd need to swap the b7g for a b9g base which could be awkward as they're riveted to the chassis. I'm sure my dad's drill press could take care of that - in fact I was thinking of drilling them out and remounting with screws and rubber spacers for the 6au6s at the very least. I also saw a similarly modified akai/roberts machine that used the spare b7g for another valve.  I forget what it was (and its purpose) but I believe it was one half of something like a 12at7 in a single tube.

Cheers
Lee

B9A is also bigger which may be a problem. Alternatively you could use the 6AR5 as the output stage and use the 6BQ5/EL84 socket for the 6922.

Mounting the first stage on rubber mounts is definitely a good idea. Probably not necessary for the rest.

Cheers

ian

 

[untune]

Are you talking mains transformers here  or audio ones?

Sorry I was referring to the mains transformer here. I've seen some models with the mains transformer shrouded, I wasn't sure if it was made like that or added afterwards

B9A is also bigger which may be a problem. Alternatively you could use the 6AR5 as the output stage and use the 6BQ5/EL84 socket for the 6922.

That's a good point. I have managed to get hold of an akai m8 which gives me two more channels to modify at some point in the future - tube compliment is different (ef86 and 12ad7 + el84 per channel) and I've been pondering what to do with them. One idea was a valve based two channel compressor, or perhaps a summing mixer/line level amplifier, or a mix of both. Swapping the el84s for 6922s would free up heater headroom on the transformers and give the option to put another valve in place of the 6ar5. It might be an idea to incorporate your compressor into that project and keep this one purely a preamp... Oh the options

Cheers
Lee

 

[ruffrecords]

Sorry I was referring to the mains transformer here. I've seen some models with the mains transformer shrouded, I wasn't sure if it was made like that or added afterwards
Cheers
Lee

The come both ways. You often get bare mains transformers with fixing holes in each corner. Sometimes a square hole is made in the chassis and the transformer is laid down so the bobbin pokes through the hole. The transformer is fixed to the chassis using the four holes. Sometime an additional shroud is used on top of the chassis, other times the transformer stand up and has a shroud both sides. They are definitely an optional add on extra. Whether they serve any useful purpose other than keeping fingers away from possible dangerous voltages is debatable. They may provide some additional screening but I do not know how effective it is.

Cheers

Ian

 

[untune]

The come both ways. You often get bare mains transformers with fixing holes in each corner. Sometimes a square hole is made in the chassis and the transformer is laid down so the bobbin pokes through the hole. The transformer is fixed to the chassis using the four holes. Sometime an additional shroud is used on top of the chassis, other times the transformer stand up and has a shroud both sides. They are definitely an optional add on extra. Whether they serve any useful purpose other than keeping fingers away from possible dangerous voltages is debatable. They may provide some additional screening but I do not know how effective it is.

Cheers

Ian

I thought so; in this case it's as you say laid through a square hole in the chassis but rather than the four corner holes, there are a pair of mid spaced holes through which it's fixed right through with two long bolts from the top and into the chassis at the bottom.

I always thought they were a little more aethetically pleasing in the least to hide the windings, plus as you say, safer if the transformer is exposed or mounted outside.

Can't seem to find them anywhere though. Hard enough to find the right term to describe them. I might fire an email to a winder - come to think of it, Danbury in the UK have just ceased trading due to retirement and are selling the winding equipment off. Maybe they have a few lying around

Cheers
Lee

 

[untune]

A bit of searching turned up the correct term for the cover - 'end bell'!

 

[untune]

Started tearing everything down over the weekend - taken longer than expected since all of the tags on the tagboard were folded over and pressed against component leads, then soldered... had to separate them with heat and a very sharp blade!

Alas, other than finishing that, it's stripped down and ready to put together; I've attached my final 'build' schematic. If anyone spots anything odd or out of place, please let me know. I'm not certain the grid stopper resistors are neccessary for triode 6au6's, at least at stage 1. But I've stuck them in just in case. I also forgot to check whether the triode strapped stage 2 and 3 valves need small resistors between screen and plate.

Since the power transformer was wired to 100v rather than 117v on the primary, I'll keep it like that. I'll use a 240-110v adapter for testing, and eventually I might put the other step down mains transformer from the Akai, complete with selector, into an external box.

I've left the VU on the schematic for the time being - the meter I sourced would need to be driven by another valve and while I do have that b7g base spare, I'm not sure I can grasp how it works!

Cheers
Lee

 

[ruffrecords]

You need a 470k  resistor form V2 grid to 0V else their may be no DC path from grid to 0V when you use the line input. Also the VTB2291 needs to be ac coupled to the EL84 anode. You cannot pass DC through it.

Cheers

Ian

 

[untune]

You need a 470k  resistor form V2 grid to 0V else their may be no DC path from grid to 0V when you use the line input. Also the VTB2291 needs to be ac coupled to the EL84 anode. You cannot pass DC through it.

Cheers

Ian

Thanks Ian With V2 grid, should R3 (500K) not cover that duty?

And noted, thanks - I thought it was strange that there was no cap to block DC at the output but again I think I misunderstood. Would say a 0.1uF cap between the 5K1 and the VTB2291 suffice? Trying to think what suitable caps I might have stashed away somewhere...

EDIT 0.1uF is too low I'm thinking... more like 4.7uF? I assume in this configuration I'll be wiring the VTB2291 for 9600R and not 2400R?

Cheers
Lee

 

[ruffrecords]

Yup, R3 does it. I missed that.

4.7uF would be a good value to use. Make sure it is rated to at least the supply voltage. A film cap is best.

Cheers

Ian

 

[untune]

Yup, R3 does it. I missed that.

4.7uF would be a good value to use. Make sure it is rated to at least the supply voltage. A film cap is best.

Cheers

Ian

I'm going to move that input closer to the second valve on the schem, it's a bit confusing the way it is

Thanks for that - some of these higher value film caps are extraordinarily expensive... £100+ in some cases 

Cheers
Lee

 

[untune]

I updated the schematic to clarify things a little bit. After looking at it more closely, I was just curious about one thing regarding the EL84 stage; mine has one side of the OT primary connected to the HT, and the other side goes through the cap/resistor and into the connected plate/grid. If I was to connect things like the previous triode stages, (i.e. plate/grid direct to HT through the 5K1 resistor, and then plate/grid to coupling cap to OT with the other side of the primary to earth) how would that affect the operation? I'm just curious about how it operates since I took that from the original schematic and didn't give it any thought.

Cheers
Lee

 

[ruffrecords]

If you are using the VTB2291 there is no connection from it to the HT. The 5K1 goes from the EL84 anode direct to the HT just like in the other stages. The 4.7uF goes from the anode to one side of the 2291 primary and the other primary connection goes to 0V.

Cheers

Ian