• WARNING: Tube/Valve amplifiers use potentially LETHAL HIGH VOLTAGES.
    Building, troubleshooting and testing of these amplifiers should only be
    performed by someone who is thoroughly familiar with
    the safety precautions around high voltages.

Request for suggestions. Akai M8 mono amp schematic validation.

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Aloha,

My electrical engineering experience stretches way back to last week Sunday when I happened to find a Akai M8 out on these beautiful amplifier-rich streets of San Francisco (HK, Marantz, now Akai found). Turns out, with the help of these forums and some elbow grease i have arrived at the following circuit physically, I have done all of the circuitry reduction noticeable in this schematic and am asking for advice for changes and assurance i haven't done something whacked.

AkaiM8.jpg


The amp sounds good at volume but there is a audible buzz @ 0-gain using the only headphones i have (Fostex orthodynamics, these are high-impedance drivers), and I intend to recap the amps with the following;

1x Orange Drop 715 Series .001µF / 600V
1x Orange Drop 715 Series .010µF / 600V
1x Orange Drop 715 Series .022µF / 600V
2x Sprague Atom 25µF / 25V
3x Sprague Atom 20µF / 350V

Very much appreciated! :geezer:
 
Last edited:
Hi,

I started a similar thread two weeks ago,

http://www.diyaudio.com/forums/tubes-valves/213857-need-help-modifying-roberts-770-monoblocks.html

In my case the open loop response is not flat and requires Negative feedback from the secondary. In the case of pentode connected it sounds much better with the feedback connection. I tested different values for the pentode feedback and a resistor between 50k and 60k worked best.

Definitely replace the caps, the non oil caps were bad and affected the bias of the second half of the 12ad7.

I did not try with headphones on mine, but with speakers it is silent. I only tested with the built in speakers and vintage alnico 4 inch drivers i had on the bench at the time.
 
I have been recommended to to increase the pi filter capacitors to 40uF, the cathode coupling capacitors to .1uF and the cathode bypass capacitors to 100uF. These should decrease ripple, and increase bass responce, respectively.

Im also very interested to know if i will need some sort of input coupling capacitor, or will it be ok with the resistor-only input currently on here?
 
Well, I'm doing my best to get acquainted with tube amps but try as i might I have not found the right resource to solve this puzzle. I have been able to model the circuit in LTSpice and am getting some very disturbing output. First off, the power supply output (on this circuit only, with others it tests great) the output appears as a sawtooth wave with a delta of 5-6 volts. There is something garbaging up the works here that i cannot see.

The attached image is the Spice model i'm working with.

Any assistance would be appreciated.
 

Attachments

  • circuit.jpg
    circuit.jpg
    111.7 KB · Views: 379
Well I've spent some time reading the forums and various sources of wisdom and have arrived at this final circuit which I expect to perform very well. Again, if anyone has any suggestions for alteration I'd love to hear your advice.

The power caps will be GENTEQ GEM III motor runs.
Cathode bypass caps will be Kemet T262 solid tantalums.
Coupling caps will be Russian teflons.
Resistors will be RCD wirewounds.
I plan to invest in two Jensen 4:1 step-down input transformers so I may use a variety of sources. (My DAC outputs 2.1VRMS)
I also plan to invest in a EMI filtering power plug, a Delta 06EK3.

Thats all for now.
 

Attachments

  • schematic.jpg
    schematic.jpg
    94.7 KB · Views: 359
Last edited:
Dave,

I'm taking your advice, I realize that the gain is definitely too high for anything but phono cartridges, even then..

So i'm considering doing away with the 12ax7 entirely and using the EF86/6267 tube in triode mode as the preamp.

Seeing the Phillips specification sheet makes me wonder; rather than insert a volume pot into the signal path, since voltage and current output are directly related to the supply voltage, a potentiometer regulating this supply would be effectively equivalent. I'm unsure about how this would affect performance.

For reference, here's a chart of the relevant EF86 triode characteristics
 

Attachments

  • ef86.jpg
    ef86.jpg
    52 KB · Views: 353
The M8 chassis has 4x octal sockets, so i figure i can build it up according to the following schematic. (I enjoy the concept of transformers and their simplicity, plus relatively inexpensive Edcor products fit in each position, so i figure why the hell not - i think i can make everything fit.)

The tubes are;

2x EF86 (triode mode)
2x EL84 (triode mode)

for a total of about 5 watts
 

Attachments

  • schemwiw.jpg
    schemwiw.jpg
    51.4 KB · Views: 316
The push pull design is simply appealing. I read that distortion decreases and low volume power would be more readily available, both things I am interested in attaining. Just the output transformes are replaced, the input and interstage transformers are added. The power transformers should hold up, they are rated for 60MA.

6v4 - ~1.5ma max
2x EF86 ~2.5ma max each
2x EL84 ~25ma max each

That is pushing the limits of the power transformer, but I do not intend to be running at full volume ever.
 
frugal-phile™
Joined 2001
Paid Member
If you want to do a push-pull amp, then you need to start with a different donor.

A pair of push-pull EL84 draw ~ 80 mA.

A pp will decrease 2nd order harmonics below the 3rd, not necessarily a good thing, and the output transformer is then running thru the kink in the magnetic curve, increasing distortion levels at low power.

If you are using the M8 as a donor, you have the means to build a mono SE amp. If you want a PP amp, time to start from scratch... a shame, because tarted up, the M8 can be a fantastic little amp.

dave
 
The one i have intact sounds great indeed, and though the gain is pretty high and there is some noticeable distortion and granularity from the unreplaced capacitors and components, I agree fully that it has hooked me on tube amplifiers. Thus the reason for the last weeks of study and postulation for what i want to achieve and what through my studying has seemed most idealistic.

The phillips datasheets, which have been absolutely the best help in decision making, have the following data (attached)

At the rectified voltages from the 250V Main, which should be between 300-250v after the initial CLC filter, these datasheets point in the direction of a anode current between 20-26ma.

The top part of this clipped datasheet depicts the triode connected characteristics (which i'm interested in).
 

Attachments

  • el84.jpg
    el84.jpg
    106.4 KB · Views: 315
Last edited:
The internet still amazes me.
Thanks to forums like these and the abundant audio related information sites I have learned a plenty over the last two months without touching and idiots guide manuals (though I have ONE on the way, i dont suppose it will change this schematic very drastically).
Dave/Planet10, thank you for noticing how flawed my initial design way, you must have laughed seeing the first circuit I posted. And for seeing me through my variations on the theme.

final.png



Plate choke and interstage, and my reasons for using them;


Plate choke to; maximize output signal swing, linearize the EF86 loadline.

2:1 Interstage to allow; stepped-down input impedance (4x lower for 2:1 step-down), full bandwidth transmission of ~40VRMS stepped-down to ~20VRMS, phase splitting!

The plate choke on the EF86 triode will be 120 Henries.
Since the EF86 will have a internal resistance of ~16k, the following calculation solves for a -3db rolloff at frequency X.
X = 16,000 / (2 * pi * 120) = 21Hz

It will be composed of two series-attached Hammond 155C chokes and I'll be taking the following gentleman's advice. (from here)
3rd, you can mount the chokes on standoffs, bottom to bottom and wire the chokes with one winding backwards. This makes the pair of chokes hum buckers. In my test setup the pair of chokes picked up 20dB less hum than a single choke.
Additionally, the choke's series capacitance will be reduced by half.

My decision to not run B+ through the primary of the interstage was simply because I cant find a suitable transformer with a primary impedance of 120H+ to allow for the rolloff that these chokes will supply.

Which brings me to the one gripe I have with the circuit, the coupling capacitor between the EF86 anode and the interstage transformer. Blocking DC on the interstage seems a good thing; but how will this play out in terms of the resonance between the cap both the plate choke and interstage primary?

I'm aware of the equation for determining the resonant frequency of an LC circuit;
f = 1/2pi*sqrt(LC)
For the 120H choke, i'll have a 45Hz resonance with the .1uf cap in the schematic (not my final choice)

If the interstage primary is about 25H (likely if I choose an Edcor), the 20Hz resonance occurs with 2.5uf cap. Pointedly, i should increase the inductance or increase the capacitance, eh? Thats a big cap for signal though...


(Any recommendations for solid interstage transformers would be helpful)
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.