Prasi very kindly posted a version of the Store's speaker protection board using solid state relays (and in a slightly smaller package). See post #708: https://www.diyaudio.com/community/...ctor-board-set-v3.247279/page-36#post-7543847
Thanks folks. I'll look into these. I already have the store board winging its way to me, but it's good to know I have other options.
I have to say I've seen the Neurochrome unit before, and while I'm sure it's very good, I'm not a fan of just plugging in a pre-populated SMD board - just feels like I've cheated.
I've been unclear. I should say I have this feeling about all pre-populated boards, simply because what I find enjoyable about DIY is doing everything, as much as possible, myself. Even if it's not the best implementation. I know you're aware of this kind of thing because you've said yourself the LM3886 Done Right is, and always will be, through hole. That market exists.
I have no doubt the component you've chosen is the best performing, and a pre-populated board is the best way of providing that performance to the customer. No doubt at all. I'm not intending to insult your work.
I just prefer soldering all the bits where I can. Personally.
I didn't perceive any insult so no worries there. I do understand that some enjoy the build process from start to finish while others are, frankly, tired of soldering everything. It's just unfortunate that those who wish to complete every solder joint themselves will likely have to give up some performance in exchange for that. I plan to cater to both segments for as long as it's practical.
Tom
Tom
The ones I use have tiny little exposed legs and I suppose could be hand soldered but a hot plate and solder paste would be needed for the wide drain electrode. I never thought about offering the PCBs as bare before because I assumed people just wanted these to work as a piece of gear.
The above is the latest version but the MOSFET footprint has a large pad underneath the 4 “pins” on one side. This is needed for thermal conduction for cooling.
Here is the PCB for v1 (which works fine as long as amp is not bridged).
Last edited:
24 volts DC can be reduced to 12 volts with a 7812 regulator. That would be the way I would look at.
For AC you need 9 volt, not 18 volt. 9 volts AC * 1.41 gives approx 12 volt DC (rms voltage * 1.41 less the 0.6 of the rectifier diode).
Thank you for that detailed breakdown of what's involved - I can certainly see why you offer the board complete. Definitely beyond my skillset at the minute. You're probably right that most (if not all) people want something that just works. I think I'm just passing the time more than anything, so the more I have to do myself, the more I keep myself busy.
Can I use +/- 12VDC to power the store's DC protector board? I plan to use a Micro Audio Cobra S2 for my Wolverine build.
Actually a bit confused by this advice. I've been making my way through both this thread and the previous V2 thread, noting down as much as I can (16 pages of notes for this tiny little board, if you can believe it). The advice for supply to the board, again and again, from the original V1 build document onwards, says that your supply to the board must be roughly double the coil voltage of the relays chosen because they're in series, and that because the AC is half rectified, the circuit is designed such that there shouldn't really be a great difference between AC or DC supply at the same voltage. In that way 24VDC can be used with two 12V relays, as can 24VAC.
Some have reported the board turning 24VAC into about 25.6VDC, but none have gone on to say this caused issues. It seems like 18VAC also works, from what I've read.
I can't claim to be knowledgeable, but that's the gist of what I've read so far.
If your relays have 12 volt coils then yes, 24 volt DC as a supply is correct. Its not very critical but if higher than that then the relay coils will dissipate more power (heat) than they should. If to low then they will not pull in smartly.
I mentioned the 5 volt option because of you mentioning:
So you have to set the PSU voltage around the relays you intend using.
You need to understand the relationship between an AC voltage from a transformer and a DC supply voltage. The DC value is what it says on the tin, so 24 volts DC is just that. An AC value is an RMS value...
So a transformer that is '18 volts AC' has a voltage output that is a sine wave in shape and that starts at zero, goes up to a positive peak value, then falls back through zero and toward a negative peak value.
The '18 volts AC' is the rms value. The peak is 18 multiplied by root 2 (1.414) which gives 25.4 volts. That is the voltage at tip of the positive and negative point of the sine and that is the voltage that rectifying the AC voltage will produce as a DC value. The rectifier conducts and charges the reservoir cap after the diode to the peak value of the AC.
https://circuitdigest.com/tutorial/ac-circuit-theory-peak-average-and-rms-values
The board runs on a single rail supply so you would use just the positive 12 volt rail to power it and you would use 5 volt relays.
You can not connect it between +12 and -12 (to try and get 24 volts) because the ground reference for the board would then be incorrect as it would be referenced to -12 volts.
Thanks for that, @Mooly. That was roughly my understanding (albeit just the VAC RMS*1.41 bit), but then I got to a post from @prasi that linked to the original build guide by JoJo (#218 in this thread), and it said the following, which befuddled me a bit:
"Choosing your Relays
Coil ratings of relays K1 and K2 must be chosen depending on the AC voltage
that you have available. The coils of the relays are in series, so you add them to
get the voltage rating requirement to power up the project.
For example, if your power transformer has a spare 9VAC or 12VAC secondary
winding, choose 6V relay coils for K1 and K2. Or if the spare secondary winding
is 18VAC to 24VAC then a 12V relay coil must be used for K1 and K2."
I puzzled over this for a while, as it did seem the AC ratings were high. Later someone posted an explanation to do with the rectification, and, not knowing how much they knew, nor seeing anyone contradict them, I sort of went with it as correct. Clearly it wasn't, so thank you again for setting me straight.
I had the presumption to think, in the absence of anyone else doing it, I could gather enough info to help update the original build guide to a usable state, but, as a newbie, it is proving challenging.
I'm learning things, however. 🙂
This might explain it better.
The input voltage is 18 volts rms. Look how the AC voltage rises to a value of 18*1.414 and also falls to below zero to that value. So 18 volts rms is 25.45 volts peak and so 50.90 volts peak to peak (top to bottom).
The rectifier diode only allows the AC part of the voltage that rises above zero to pass and the reservoir cap (C6) charges to the peak voltage which is 25.45 volts. In practice the diode drops around 0.6 volts but we can ignore that.
The DC voltage has 'ripple' on it and that is because when the AC voltage falls away the circuit gets its power from the cap and the voltage on the cap falls. Notice how the red voltage line isn't straight but falls a a little between each peak.
When the next positive half cycle comes along the diode tops up the cap again once the incoming voltage exceeds the voltage remaining on the cap. That happens every 20ms for 50 Hz mains.
The input voltage is 18 volts rms. Look how the AC voltage rises to a value of 18*1.414 and also falls to below zero to that value. So 18 volts rms is 25.45 volts peak and so 50.90 volts peak to peak (top to bottom).
The rectifier diode only allows the AC part of the voltage that rises above zero to pass and the reservoir cap (C6) charges to the peak voltage which is 25.45 volts. In practice the diode drops around 0.6 volts but we can ignore that.
The DC voltage has 'ripple' on it and that is because when the AC voltage falls away the circuit gets its power from the cap and the voltage on the cap falls. Notice how the red voltage line isn't straight but falls a a little between each peak.
When the next positive half cycle comes along the diode tops up the cap again once the incoming voltage exceeds the voltage remaining on the cap. That happens every 20ms for 50 Hz mains.
I really need to learn how to use LTspice - that seems a really handy thing.
The original BOM specs C6 as 220uF - 470uF. Does going larger improve the ripple any? Truth is I'm probably going to supply DC to the board from the F5M cap board (if that's doable), but it would be good to know about the AC option if I go that direction.
Thanks again for the help.
The original BOM specs C6 as 220uF - 470uF. Does going larger improve the ripple any? Truth is I'm probably going to supply DC to the board from the F5M cap board (if that's doable), but it would be good to know about the AC option if I go that direction.
Thanks again for the help.
A bigger cap will reduce ripple but that isn't always a great idea in every case.
Using AC and a 'small' cap means that as soon as the mains is disconnected the current draw of the circuit will collapse very quickly and that is exactly what you want here to ensure the relays drop out 'instantly'.
Using AC and a 'small' cap means that as soon as the mains is disconnected the current draw of the circuit will collapse very quickly and that is exactly what you want here to ensure the relays drop out 'instantly'.
That's really very helpful. Thank you.
So, powering it from the massive amplifier DC supply is a bad idea because that'll take forever for the board to lose power? The AC option is starting to sound like the way to go.