FireFly mini-amp build report

[klooon]

Here is a small build report for the FireFly mini-amp. I was inspired by this site:
https://diyguitarfreak.wordpress.com...ecc99-miniamp/

and used this schematic:
https://diyguitarfreak.files.wordpre...schematic1.jpg

I have added a small 8Ohm 2Watt speaker (28mm diameter) for fun and protection. Further, I have added an attenuator (resistive load), for 40% reduction of the output power...

Here is the layout I used for my Hammond 1590XX enclosure (light gray).


And here are some pictures




Sounds really good! First had a problem with hum and buzz, but that was solved by taking shielded wire to V1-pin7 (grid). Now it's really quiet. Nice little amp!

[guitician]

What is the frequency response of that tiny speaker? It looks like you could have got 4 to fit, and got more bass. Nice build! Any sound demo?

[klooon]

What is the frequency response of that tiny speaker? It looks like you could have got 4 to fit, and got more bass. Nice build! Any sound demo?

Here is some more info on the speaker:

I did not put much thought to it, other than that it was the smallest 2W speaker available (and cheap... 2.50 euro). The 4x1"would be a nice option

[klooon]

Here are some measurements:
- wall voltage: 227VAC
- B+ (@A): 256.6 VDC
- B+ (@B): 249.2 VDC (not much voltage drop...)
- 'power tube' ecc99: plate voltage: 252.2VDC; V_Rcathode = 7.70VDC; I_Rcathode = 7.7/220 = 0.035A; dissipated power of the two tube triodes: (252.2 - 7.7) * 0.035 = 8.6W; per tube half this is 4.3W which is 86% of the the 5W maximum rating...

Would this be acceptable or is it better to increase the cathode resistor?

Further, I have added a conjunctive filter by means of a 0.0022uF (630V) capacitor. Just like the original FireFly circuit (link). This takes away some harshness (in the overdrive) and especially with the tiny internal speaker...

[guitician]

Wow, not much bass to be had with that one... you would need about a dozen with a folded horn to bring up that roll off. That filter to knock down the high-end may make it sound more like a guitar speaker. Put the amp in a big cardboard box, maybe it will sound bigger.

The 125A has a max Pri. rating of .025A DC and your at .035A... you may want to bring it down to be safe.

[klooon]

Wow, not much bass to be had with that one... you would need about a dozen with a folded horn to bring up that roll off. That filter to knock down the high-end may make it sound more like a guitar speaker. Put the amp in a big cardboard box, maybe it will sound bigger.

Haha. I put a wire to a 4x12 or 2x12 and that sounds big Is there a tiny speaker speaker available with a better response?? What should I look at?

The 125A has a max Pri. rating of .025A DC and your at .035A... you may want to bring it down to be safe.

Never looked at this rating before, just looked at tube specifications. OK, I should stay under .025A then. I am now using a 220R, I will try a 330R...

But 0.025A would mean about 60% of the max. dissipation. Which should be fine? Better use another transformer like the Hammond 125C with this design for instance?

Or alternatively, is it 'better'/an idea to bring the voltages down a bit with a 'sag' resistor between the rectifier and the first filter cap? I tried 1.2k but it brought down the voltages too much (<200VDC instead of the 256VDC)... If this is a way to go, what value resistor should I try/stock?

[klooon]

I have changed the cathode bias resistor from a 220R to a 330R and the current draw is now 25mA instead of 35mA for the two tube halves together. This is now within spec of the 125A output transformer.
Some more values: B+(A)= 263VDC; B+(B)= 256 VDC; plate voltage= 261VDC; V(330R) = 8.5V; 8.5/330R=25mA; power: (261-8.5) * 0.025 = 6.3W and per tube halve: 3.15W which is about 63%.


Further, I have tried a 220R 'sag' resistor between the rectifier and the B+(A) -- I didn't have a different value at hand... (still with the 330R cathode bias resistor). Recifier output voltage is about 255VDC and B+(A) = 250VDC. Plate voltage has come down a bit to 248VDC and the voltage over the bias/cathode resistor is now 8.0VDC instead of 8.5VDC. The current draw is now 24mA (=8/330) and the resulting dissipation power per tube halve 2.9W (58%). Not so much difference...

I put the 220R a bit in the air and now the noise/hum is back...

BTW. The 220R and 330R I used are 5W resistors...

[guitician]

You could try removing some turns on that toroid transformer. Either the pri/sec is wound on the outside. If it's the secondary you can unwind a few turns to bring down the voltage.

[klooon]

I have tried different 'sag' resistors (although I understood that this resistor in class A doesn't sag...), of values 330R and 1.2k and this doesn't bring down the current draw so much without dropping the bias too much (I don't want to bias much less than 60%). With the 1.2k and a bias resistor of 330R, the total current draw is 20mA, but the bias is 40%. I have ordered a 390R bias resistor, that I will try without a sag resistor. This should bring down the current draw below the 25mA (which is the max. for the 125A OT) to be a bit safer and should bring the bias between 50 and 60% which is still acceptable (??).

[guitician]

Have you looked into replacing the OT...12ax7-ecc99 tube amp
Biasing the PP ECC99 at 70% max power means going over the 125A's current rating. Unless you can increase the B+ to keep the current lower. A 5 watt trans. may be easier.

[klooon]

Have you looked into replacing the OT...12ax7-ecc99 tube amp
Biasing the PP ECC99 at 70% max power means going over the 125A's current rating. Unless you can increase the B+ to keep the current lower. A 5 watt trans. may be easier.

Thanks. Unfortunately, I have bought two 125A OTs for two FireFly's, so I will try to stick to them - if possible. So I will first go this route. I have tried to lower the B+ by the 'sag' resistors. It drops the B+ quite a bit, but with a targeted slightly less current draw, this means much colder bias (ca. 40% of 5W). I figured out that only with a slightly higher bias resistor value I can lower the current draw (target: under 25mA in total for the EC999) whilst keeping the bias not too cold. Therefore I want to try a 390R bias/cathode resistor. Currently, the bias is at 63% of 5W and the current draw at 25mA and the sound is really great. I know 25mA is the specified maximum 'bias current draw' and - although on the 'edge' - maybe it's fine...

[ThisLifeILead]

thats a tight little beast!!

[eschertron]

...and the sound is really great.

Have you posted any sound clips? I'd like to hear it

[klooon]

OK, I finally got the new 390 Ohm bias resistor in. Now the B+ is 268VDC (and 260VDC at the second filter node). The plate voltage is at 265.8VDC. The voltage drop over the bias resistor of 390R is 8.9VDC. The current over the both tube halves is (8.9/390=) 22.8mA, which is nicely within the specification of the Hammond 125A (25mA max). And the bias is at 59% of the 5Watt maximum rating of the ecc99 tube. --> [(Vplate-Vbias)* current / 2] => 2.94 W.

All fine...now...

[klooon]

Further, I found that the attenuator (to 40%) wasn't bringing the power down enough (difference was hardly noticable), so I have adapted the attenuator resistor values to bring it down a bit more. R1 is now 15 Ohm and R2 is now 10 Ohm. This leads to 20% of the original power. This is quite usable.