Pioneer HPM-100 circuit questions. (Super tweeter and transformer)

[Audiotexan]

Not being up on the theory side... I figured I'd bring what I've found thus far here.

My current situation is this: I have a pair of HPM-100's that has no response from the super tweeters.

I've got 3 replacements from ebay (two are a matched pair, 1 for spare), which upon arrival all 3 were tested (confirmed good) via my Wavetek audio generator (using manual sweep).

I try installing the one oddball into a cab and still nothing.

So I dig into that crossover, and desolder the component junctions as needed to test the 4 points individually.

- The 10 ohm/5 watt tests good (10.8 ohm).

- No way to test the .1 mh choke (lack of LCR meter ATM) -but it's got connection according to the old Fluke 87. (.1ohm or so IIRC)

- No real way to test the .15uf (again, lack of LCR)
But did test it like an electrolytic as I've done in past. After shorting, I did see charge ramping up...only it went infinite!
I figure it should have stopped around 10meg-20meg give or take, but wasn't sure, due to it being non-polarized.)
*The cap is seeming highly suspect at this point.

But when I take another N.P. 100nf (closest to 150nf I could find) and test it using the same method as above, I get the same result. So I'm left without a definitive answer as yet. (I *really* need to get a good LCR...it's just hard to spend so much when they're so rarely used. >.<
But when you need them, nothing else will do!)

- The transformer. Here is primarily why I wanted to present all this for your input. (Though of course, I appreciate your input on any/all of it!)

Once I got to the transformer, I measured the secondary (speaker leads attached) and got approx. 44-45 ohms. Cool.
When I measured the primary, I got 0.8 ohms in ckt. Disco'd both leads of primary from the rest of ckt. Still get .6 - .7 ohms reading. I was all "Voila!" Bad transformer!
Immediately followed by "Damn. Bad transformer..."

So I got to searching a bit, and found this link Help! - Specs for Pioneer HPM-150 ST Transformer - AudioKarma.org Home Audio Stereo Discussion Forums
which goes into folks taking some measurements...and then I feel a touch of possible relief.

It seems that .6 ohms on the primary might not be the "shorted" condition I assumed it was. And it's close enough to what another member there measured (.4 ohm primary, 44 ohm secondary) on their HPM-100's with the STV-013 transformer (which is supposed to be a 1:8 transformer).

..they need a high voltage, but almost no current, due to their piezoelectric nature.

I was thinking they were more of electro-static?
Or am I again confused? Either way...I confuse the two since it's rare I deal with them.

My transformer question is, they brought up the turns ratio vs impedance ratio, so I'm hoping that I'm close enough to the .4 primary (as mine is reading .6), or am I out of spec? NO clue on how to go about the math on this one. But given the ultra low 'tolerances' already, I have to wonder if *in this case* that 2 tenths difference is enough to warrant concern? Or am I overthinking it, because of not knowing the math in this case?

The first paragraph in Post 7 by GordonW is where the math lost me. But it also made me think that my transformer may actually be good.

Either way, I'm glad I didn't gut it to see if I could find a shorted winding before I went looking for info!



Regards,
AT

[Audiotexan]

hrmm... no one has a good link on this guy's conversion process??

Yep... if the wire gauge is similar on primary and secondary, 0.4:44 ohms equals about a 1:110 impedance ratio. That would imply about a 1:10:48 turns ratio, roughly. Extrapolating from that, it looks like it would be 1:8 turns ratio... or 1:64 impedance...

0.4:44 = 1:110 impedance I get..
But how does he get turns ratio from that?

And then, how does he turn the 1:110 into having a third group of numbers? 1:10:48 (Where's the 48 come in/from?)

So is he rounding down from 1:10 to 1:8?? Or are his "extrapolations" assumed to be understood?

How does he go from: "0.4:44 ohms equals about a 1:110 impedance ratio" to having the last line read "1:64 impedance"??

Someone please enlighten me on all this.

[nickb]

hrmm... no one has a good link on this guy's conversion process??



0.4:44 = 1:110 impedance I get..
But how does he get turns ratio from that?

And then, how does he turn the 1:110 into having a third group of numbers? 1:10:48 (Where's the 48 come in/from?)

So is he rounding down from 1:10 to 1:8?? Or are his "extrapolations" assumed to be understood?

How does he go from: "0.4:44 ohms equals about a 1:110 impedance ratio" to having the last line read "1:64 impedance"??

Someone please enlighten me on all this.

Ignore it. The DC resistance of the windings has nothing to do with the turns ratio.

You need to measure the voltages in order to determine the ratio. Given this is a fairly high frequency transformer I think you should drive the crossover with say 10Khz via the power amplifier and use your meter on an AC range across the primary and then again across the secondary. The ratio of those voltages is the turns ratio. The snag is that not all meters will work up to 10khz as they are often limited to 50-60Hz. I 'm banking on the meter working not necessarily reading accurately but remaining linear so the the ratio will be correct.

Edit: Something else to consider is that these tweeters really are super high frequency so your hearing needs to be good. The high pass filter for the super tweeter in the crossover has a -3dB point of 32Khz and is -24dB at 10Khz. My old ears don't work up there! Borrow a dog

[J M Fahey]

Agree and add:
1)

- No real way to test the .15uf (again, lack of LCR)
But did test it like an electrolytic as I've done in past. After shorting, I did see charge ramping up...only it went infinite!
I figure it should have stopped around 10meg-20meg give or take, but wasn't sure, due to it being non-polarized.)
*The cap is seeming highly suspect at this point.

Quite the contrary, that means it's a very high quality cap .
Those 10/20M you measure on others actually means loss.

2)

upon arrival all 3 were tested (confirmed good) via my Wavetek audio generator (using manual sweep).

what does this mean?
You fed audio into them and heard it through the tweeter?
If so, repeat but feed audio unto transformer primary, connect secondary to tweeter and listen again.

If you heard it before (congratulations kid ) you must hear it much more now, since that transformer applies a 18dB boost

It will be much harder for the generator to drive it, so you should first drive an amp (don't overdo it, just enough for it to put out, say, 3 or 4 V RMS) and then the transformer and tweeter.

Post results.

3) these are piezoelectric tweeters, but since that word got an undeserved bad reputation they made up another.

If standard Piezos worked exactly the same but cost, say, U$200 each, there would be Cathedrals built to worship them, but who respects $2 parts?
Even though they are flat to 24/30 kHz.

[Audiotexan]

Ignore it. The DC resistance of the windings has nothing to do with the turns ratio.

Much appreciated. I was completely lost on trying to follow him, and felt quite foolish as a result.
Part of what he was saying would make sense, and then he went off into numbers and I was like "...erm...what? How? Do huh??!?!"

So, I figured it was better to bring it here, and ask those I know/trust.

You need to measure the voltages in order to determine the ratio. Given this is a fairly high frequency transformer I think you should drive the crossover with say 10Khz via the power amplifier and use your meter on an AC range across the primary and then again across the secondary. The ratio of those voltages is the turns ratio. The snag is that not all meters will work up to 10khz as they are often limited to 50-60Hz. I 'm banking on the meter working not necessarily reading accurately but remaining linear so the the ratio will be correct.

Ahh good point about the meter I wouldn't have thought of.
TY for that as well!

So it'd be easier to see scoped it if/while it's still in circuit. Got it.

Edit: Something else to consider is that these tweeters really are super high frequency so your hearing needs to be good. The high pass filter for the super tweeter in the crossover has a -3dB point of 32Khz and is -24dB at 10Khz. My old ears don't work up there! Borrow a dog

Ok. Do you have a suggested value cap to boost the 8-15k range then? (Unsure how many db would be appropriate, but -24 sounds like it's WAY too steep for what should be coming out of there. I'd think it would've been set more to -6 or maybe even -9 @ 18k (so it's not overbearing or harsh), and maybe -3db to -6db around 8-10k, to pickup where the regular tweeter leaves off)

Agree and add:
1)
Quite the contrary, that means it's a very high quality cap .
Those 10/20M you measure on others actually means loss.

Ok. I forgot something along the way, due to the fact that the "number" (eg: 10mg, 20mg, etc) would change on an electrolytic with regard to the amount of storage it holds.
Makes sense why a N.P. wouldn't "top out" since it doesn't hold a charge like electrolytics do.
Amazing what a clear head (or a smack upside it to jar the cobwebs loose) will do. Thank you both for replying!!!

The whole "using the resistance setting to measure capacitance" in a pinch (and watching the 4000 count strip below the primary counter to see how fast it charges) was a trick my first boss showed me to be able to 'rough check' caps if you didn't have an LCR meter handy. So I've come to rely on it more than I should.

Couple that with so many OEM's cheaping out, and using electrolytics where they should be using proper coupling caps simply because the 'lytics are cheaper doesn't help. But at least there, it's usually obvious for me and I make the implied transistion without issue.

2)
what does this mean?
You fed audio into them and heard it through the tweeter?

Correct Juan! When I was 28, I could hear up to 22k with no problem. And was pretty on target with frequencies to boot...(blind testing). When I tested the replacements I received, I was struck by the fact that they'll actually reproduce down to 5-6k, but had the most emphasis up around 10-13k (with only the sig.gen. pushing a very light signal.)

Now what concerns me is:
1. either these speakers aren't capable of the range they say they are (doubtful)
2. my hearing has dropped off by an extreme amount since I could only register up to 13.5-14k (possible, but would hate to have that be true as I'm only mid-40's)
3. I didn't have sufficient voltage to push them into the upper registers (BIGGEST probability)

If so, repeat but feed audio unto transformer primary, connect secondary to tweeter and listen again.

If you heard it before (congratulations kid ) you must hear it much more now, since that transformer applies a 18dB boost

Ahhh...ok. Great call on that. Don't know why I didn't think of that before myself since I've got the transformer pulled currently. *sigh*
See? Little things like these when I'm not thinking straight are what make me love this place! lol

So taking Nick's info into account, since the transformer is adding 18db back into things, does that mean that there's a sum total of -6db attenuation ("-24dB at 10Khz"), and subsequently +15dB at the 32Khz notch ("-3dB point of 32Khz")?
<Nick: Forgive me if you already took that into account and numbers are what you stated!! I had to ask for my own clarity.>

It will be much harder for the generator to drive it, so you should first drive an amp (don't overdo it, just enough for it to put out, say, 3 or 4 V RMS) and then the transformer and tweeter.

Post results.

Will do sir!

3) these are piezoelectric tweeters, but since that word got an undeserved bad reputation they made up another.

If standard Piezos worked exactly the same but cost, say, U$200 each, there would be Cathedrals built to worship them, but who respects $2 parts?
Even though they are flat to 24/30 kHz.

*chuckles and nods* How true. Funny how some folks' opinions change when you take the same thing and charge 5-10x as much. "Oh it must really be special!!"

[Audiotexan]

Also, it just hit me....and I know you're *rarely* wrong Juan....but wouldn't that make that transformer be 24db boost? (If it is 1:8 ratio)

IE: if I hit the primary with +3, then shouldn't the secondary put out +24db?

Or am I missing something (probably) where you arrive at 18db?

[nickb]

...So it'd be easier to see scoped it if/while it's still in circuit. Got it.


...Do you have a suggested value cap to boost the 8-15k range then?

Yes, a scope would be better.

Try 0.33uf or even higher BUT these things can be finely balanced. You risk making it sound tinny at best, saturating the inductor, saturating the transformer or even damaging the super tweeter. Without the actual data for these I would leave alone unless it's for a low power short term test.

[J M Fahey]

but wouldn't that make that transformer be 24db boost? (If it is 1:8 ratio)

IE: if I hit the primary with +3, then shouldn't the secondary put out +24db?

Or am I missing something (probably) where you arrive at 18db?

As often happens, truth is simpler than myth.

While you imagine mysterious mental powers, truth is I made a rough mental estimation first, what some might call "ballpark"/ "experience" / "seat-of-the-pants reckoning" / etc. and remembering that 10:1 in voltage amounts to 20dB ; thought 8:1 would be "a little less" , which for starters is not bad.
No NASA Science here

Then immediately googled "voltage ratio to db calculator" and found:
dB Voltage Ratio Calculator

Do it yourself: fill the windows with voltage values , in this case 1V and 8V , hit "calculate" and bingo!!! : 18.1 dB

That said, now I feel a moron, because 1 minute ago, writing this, I noticed I could easily have calculated it in my mind

How?:
2:1 is 6dB , an easy to remember fact.
8 is 2*2*2 (in voltage) so 6+6+6 (in dB)=18dB .

Or to put it in a more verbal way, no pencil and paper involved:
"every voltage doubling means 6dB increase".
"Since there's "3 doublings" there's 3x6=18dB increase"

[Audiotexan]

Try 0.33uf or even higher BUT these things can be finely balanced. You risk making it sound tinny at best, saturating the inductor, saturating the transformer or even damaging the super tweeter. Without the actual data for these I would leave alone unless it's for a low power short term test.

Yeah, I do too! If the manufacturer could be sourced, that'd make things much easier. Or at least give an approximation.

The Pioneer part number for the Super Tweeter is HPA-015-01.

The Pioneer part number for the transformer is STV-013, but it looks like they did an internal conversion (instead of using the manufacturer's ID code) as I did try and find what I could using the part number, and that wasn't much. Only real thing I came up with was at PacParts.com (which was $10.32 when they last had it) but is long since NLA.

Was hoping that at least the HPA and STV was an abbreviation for the manufacturer...but it doesn't appear as such.

I can say that currently, they have the crossover point selected at 12k. I'd prefer to see that down around 8-9k.

[Audiotexan]

...I was struck by the fact that they'll actually reproduce down to 5-6k, but had the most emphasis up around 10-13k (with only the sig.gen. pushing a very light signal.)

Having retested things...they'll actually reproduce down to 3k, however, you can definitely sense that it prefers 8k-10k upwards.
I'm guessing that the resonant frequency is somewhere between 10k-12k. So I'm betting that once I can figure out whether I can drop the crossover down to 8k-10k safely (and ideally without having to change the .1mh choke) maybe it will become usable even if it's still at the 24db rolloff that Nick mentioned.

It will be much harder for the generator to drive it, so you should first drive an amp (don't overdo it, just enough for it to put out, say, 3 or 4 V RMS) and then the transformer and tweeter.

Post results.

Will do sir!

Well, I set the sig. gen. up for 1v out (for simplicity), but then when I actually put the load on, it dropped substantially. So I dialed it back to an even 250mv output @10K, and measured the secondary with the tweeter connected, and got 1.995 volts. So that seems to confirm both the 1:8 voltage ratio, as well as the 18.1db gain of the tranny. Thanks again for a great link Juan! As well, of course, for the math/explanation!!

Interesting to see that it's nothing to do with turns ratio or impedance ratio as hypothesized before on the other site(s).
(I'm so thankful for this place!! )




Also, Still need clarification/confirmation on the following to see if I'm understanding things correctly? Or if I need to take Nick's numbers as previously posted?

So taking Nick's info into account, since the transformer is adding 18db back into things, does that mean that there's a sum total of -6db attenuation ("-24dB at 10Khz"), and subsequently +15dB at the 32Khz notch ("-3dB point of 32Khz")?
<Nick: Forgive me if you already took that into account and numbers are what you stated!! I had to ask for my own clarity.>

[nickb]

The transformer gets you 18dB of voltage gain. You are confusing that with power gain I think, of which you get a little less than unity.

The power gain is what matters = gain of crossover x gain of transformer x gain of tweeter at each frequency of interest.

BTW, if you do make the 0.15uf bigger than about 0.47uF you will make the crossover filter rather peaky (resonant) and that is not likely to be a good thing. You would need to change the inductor too to avoid that. Anyway...

You original problem was that the tweeter wasn't working but now you understand things better and have checked all the components I think you've learned that it is working so..what's the problem?

[Audiotexan]

...since the transformer is adding 18db back into things, does that mean that there's a sum total of -6db attenuation ("-24dB at 10Khz"), and subsequently +15dB at the 32Khz notch ("-3dB point of 32Khz")?

Rephrased:
What I was trying to ask was, if you arrived at this via math (which I was assuming you had [or possibly a simulation program], and if so, had you already accounted for the 18db added by the transformer) or found crossover specs, or...?


Since you stated quite clearly that it's got:
-24dB attenuation at 10Khz
-3dB attenuation at 32Khz

I was curious if I still needed to adjust the numbers from there (by adding the 18db to the above) or if you had already done so.

If you had already added in (factored in) that +18db, then nothing changes from what you stated.

If that +18db wasn't part of your calculations/sim, then (if my math is right) it should be:
-6dB attenuation at 10Khz
+15dB boost (gain) at 32Khz

That's all I was trying to figure out with regard to that question. =)
I honestly don't believe at this point that the tweet in question can reproduce anything above 15k, but I'm really leary of putting anything more than a couple volts into it direct without being installed into the crossover. Perhaps that should be the next step. I don't know at this point. But I'm thinking it'd be easier to try the cap mod. (Trying to stay with the simple/obvious first, before delving headlong into things as I tend to do. So, as always, correct/remind me of the overlooked as needed)

The transformer gets you 18dB of voltage gain. You are confusing that with power gain I think, of which you get a little less than unity.

The power gain is what matters = gain of crossover x gain of transformer x gain of tweeter at each frequency of interest.

It's all academic at this point FWIW, but that's why I'm asking, I want to try and plug back in what I've forgotten because sadly, I virtually never use these maths in my day to day work, and also, it's always nice when I can learn something that I didn't know before!

Ok, so I'm guessing the Fs (since it's a closed/sealed driver the cabinet has no bearing so Fc wouldn't apply) to be approx. 10k. Given that's the case, it seems silly to cross it at 12k. (I can't 100% confirm this as I don't have a db meter. But that's what my ears tell me.) I'd think it'd be better served crossing it just under Fs (if it IS 10k) at approx. 8k, that way you get a nice 2k width above and below Fs.

BTW, if you do make the 0.15uf bigger than about 0.47uF you will make the crossover filter rather peaky (resonant) and that is not likely to be a good thing. You would need to change the inductor too to avoid that. Anyway...

Point taken, but I think it could use a 'peak' there -at least with respect to what it's got currently. Assuming for the moment that you DID factor in the transformer in your calculations (-24dB attenuation at 10Khz) it's no wonder that there's next to nothing coming out of that speaker at normal volume levels. (I never get to crank them, so, the operation at 60w-80w or even 100w doesn't matter sadly.) This would require more 'substantial' tweaking of the circuit.

If your calculations did NOT include that +18db, and it's actually -6dB attenuation at 10Khz, then only minor tweaking should have to be done, and that's why I was looking at starting with the easier point: shifting things down from the current stated crossover point of 12k to something more useful such as 8k, hopefully via the minor change in capacitor alone.

You original problem was that the tweeter wasn't working but now you understand things better and have checked all the components I think you've learned that it is working so..what's the problem?

None on that front TY both again!!!

Definitely able to confirm operation of the transformer, and the cap as they currently sit in the cab. Also confirmed that the two super tweets that were originally installed were blown.

However, they're still so rarely audible (even on cymbal-heavy music) in the mix, is what raised my questions above to start with.

Sorry if I lost ya'll by jumping ahead/around. Given the suggestion of applying signal to the tranny and speaker out of circuit, answered all of the "suggestions on testing" and since I'd already moved past figuring out what was good and bad, I jumped straight into "how to make them more usable" without saying so. Sorry for not being clearer.

[Jazz P Bass]

I am truly trying to follow all of this.
My hat is off to the math wizards.

My only question is 'why crossover the tweeter at 12K?'

Xovers that I have worked on run the gamut from 1.5K to 4K for silk dome 1" tweeter.
Maybe up to 8K for a bullet tweeter.

I would assume that it depends on the resonant frequency (FS) of the tweeter itself.

[g1]

My only question is 'why crossover the tweeter at 12K?'

This is the super tweeter, there is also a tweeter, drawing is in first post.
So my question, why do you want to lower the xover point for the super tweet?

[nickb]

In principle, your figures of -6dB and -15dB are correct if you include the transformer .

But... they don't reflect the actual system performance. It's the sound pressure level that hits you ear that matters and the system i.e. filter, transformer, tweeter, enclosure e.t.c. is much more complex than that simple model.

Hope that helps.