Electro-Harmonix Freedom Amp -- fizz on all notes

[xtian]

On the bench, a 2000s Electro-Harmonix Freedom Amp. Was sold as non-functioning, but really just needed a new lead-acid battery. Sadly, there is a low-volume, fizzy distortion on all notes, at all volumes. This issue is not the speaker.

I have emailed EHX for a schematic, but not expecting results. There's not much available via Google, either.

One odd thing. The 18v DC power adapter is meant to charge the battery only--it will not power the circuit (the amp will not function if the battery is disconnected). The LM317 is supposed to regulate the DC from the adapter, and there is a trim pot to set the voltage. However, the pot has no effect on the output voltage. Instead, the voltage out is always voltage in, minus the drop from the two diodes. Anyway, this circuit is completely separate from the rest of the amp, so I'm ignoring it for the time.

Any ideas for chasing the fizz?

[xtian]

Wow, they got right back to me!

[Jon Snell]

It will not function with the battery out as that is how the switch is designed.

[g1]

Do you see crossover distortion on the scope?

[xtian]

Do you see crossover distortion on the scope?

I'm aware. Thank you.

[g1]

I think this remark was pointed at Jon's post #3

[xtian]

Do you see crossover distortion on the scope?

No, the 1KHz sine wave output is very clean and nice looking. Even after clipping, there is no x-over notch.

[xtian]

My last report (clean sine wave) was based on no-load condition (scope clipped to output wires, but no load). This time, I connected a 4-ohm resistive load, and the output waveform looks much less nice. Does this give you any ideas?

[xtian]

Again, I contradict myself--the waveform is only distorted if the Tone and Bite controls are non-zero. If Tone and Bite are at zero (full CCW), sine wave looks perfect at all gain levels, until clipping. Still, no x-over notch seen.

[J M Fahey]

Thanks for posting, hadnīt seen any since the original ones.

LOVE the concept behind it, somewhat diminished by having only 15W out.
"Freedom" as in playing on top of mountains, coral islands or the middle of the jungle implies macho 60W or 30W plus a Celestion Greenback, both flavours available.

Excellent preamp, very poor and complex power amp, specially for meager 15W

I am quite certain you are overloading the input stage, try feeding it only 100mV RMS or, say, 300mV PP tops and recheck.

Please show a general inside picture and the speaker.

Your battery charger IC is dead, replace it.

Adjusting charging voltage with no load is meaningless, connect 100 ohm 2W resistor as load after D7 (from its cathode to ground) and *then* adjust trimmer.

The waveform you show is fine, no harsh clipping, no oscillation, clean enough for a Guitar.

Power amp design is cheesy, power transistors are working in Class B and R22/R30 barely pull a little current through bias diodes to turn them partially ON, still short of whatīs needed.
That cheap trick is whatīs used to make cheesy L358 or LM324 "usable" in Audio, they have a very similar output stage so they force a littls DC current (here 0.5 mA) through them.

But leave as-is OR replace full power module by any BTL "car chipamp", (I used TDA2005 in my version) or even a ClassD one.

[xtian]

I am quite certain you are overloading the input stage, try feeding it only 100mV RMS or, say, 300mV PP tops and recheck.

Already did. I did all the permutations, like starting with amp's Volume up high, and sine amplitude at zero, and easing it up. In all cases, as long as Tone and Bite controls are at zero, 1KHz sine wave output looks faithful to input.

Please show a general inside picture and the speaker.

Speaker is marked 4ohms.

Your battery charger IC is dead, replace it.

Just ordered LM317 and all electrolytics from mouser.

Adjusting charging voltage with no load is meaningless, connect 100 ohm 2W resistor as load after D7 (from its cathode to ground) and *then* adjust trimmer.

I trust you, but notice that schematic has a note: "14.6V ADJUST TRIM1 with S1 OFF", meaning no load.

Power amp design is cheesy, power transistors are working in Class B and R22/R30 barely pull a little current through bias diodes to turn them partially ON, still short of whatīs needed.
That cheap trick is whatīs used to make cheesy L358 or LM324 "usable" in Audio, they have a very similar output stage so they force a littls DC current (here 0.5 mA) through them.

Do you think the fizz I'm hearing CAN be improved, or is it because of what you describe?

But leave as-is OR replace full power module by any BTL "car chipamp", (I used TDA2005 in my version) or even a ClassD one.

I think this one is a flipper. Just want to get it sounding proper. Thanks, Juan! You da man!

[J M Fahey]

I trust you, but notice that schematic has a note: "14.6V ADJUST TRIM1 with S1 OFF", meaning no load.

Trust me
Those instructions are the "lazy manīs instructions"
Truth, like onions, has many layers.

* Lead Acid batteries "charged" voltage is 12.6V
Which you will measure after full charging, and letting them sit for a couple hours, no load (or a light load) Take notice of the "couple hours" delay.

Now, while charging, you need higher voltage to *push* curent into batteries.
But how much is acceptable? Pushing little takes forever charging, pushing too much is fast but if sustained dries battery liquids fast.

So standards were set long ago by the Car Industry, consumers of a zillion lead acid batteries since forever.

* car generators and alternators have a built.-in overvoltage cutoff set to 14.4V which is voltage reached with engine running, the age old electromechanical "voltage regulator" since forever, which charges them FAST. No doubt replaced today by Electronics and even microprocessors, but the old standards are still adhered to, for backwards compatibiity.
And car chipamps, VHF radios, etc. , are often (always?) rated at 14.4V supply because you DO have that voltage available with engine running.

BUT if you continuously apply 14.4V to a battery, it will have a short life.

* so next compromise was achieved: if you use a battery charger, and often forget to turn it off, they are set for 13.8V ... which is "safe".
You can forget it ON for a weekend and no big deal, reaching that voltage battery takes only a few mA, what is called "trickle charging".

Mind you, batteries either charged for a short time at 14.4V or a looong time at 13.8V will get down to 12.6V anyway, thatīs the "chemical" voltage, devoid of any charging.

* so EH instructions "should" read: "charge battery to 14.4V and unplug charger *or* set charger to 13.8V UNDER LOAD and sleep without worries"

But that means you will have to wait 12/14 hours to ensure battery is fully charged ... which Tech can wait that much?
So they went to a quick and dirty method.

My main beef with it is that they are not considering the voltage drop across the diode.
But .. everybody knows that ... 0.7V! ... or 0.65V! ... or 0.6V!
Problem is, all or none are right, depending on current through it, and with no current or a microampere (whatever the meter takes) value is undefined.

Thatīs why I like to pass a little current through diode, I trust it more.

Do you think the fizz I'm hearing CAN be improved, or is it because of what you describe?

Might be improved, but would probably involve cutting tracks, adding components ... not worth it.

[Jon Snell]

In an ideal world, a lead acid battery whether wet or glass mat needs a charge of at least 13.6volts and no more than 14.4v.
With self maintaining chargers, the current charging the battery must be no more than 10% of the A/h rating of the battery and when the voltage achieves 14.4 the charger automatically drops its charge voltage to 13.6v and stays there.
That keeps the battery full and does not over charge it or damage it.
If you leave it at 14.4volts it will boil dry in time. That is why, in the old days of mechanical voltage regulators, we had to top up the car batteries with distilled water from time to time. In the worst case, the plates would bend and short out the cells.
The fizz could be the speaker cone rubbing or low battery.

[Helmholtz]

Just wondering, the scope display shows a peak to peak voltage of 22.7V. This corresponds to 8Vrms (ignoring distortion) and a power of 16W into 4R. So far so good.
But the display reads Vrms = 15V, which is obviously wrong. I guess the scope expects you to center the curve for correct rms readout with AC coupling. Also Vavr should be zero except if you have a DCV component.

[g1]

Did it still sound fizzy with tone & bite at zero like that?

[xtian]

Did it still sound fizzy with tone & bite at zero like that?

Yes, unfortunately.