PB response and how to fix it

Alex SE

Journeyman
Joined
Oct 10, 2021
Messages
246
Location
Sweden
Tagline
Amateur
#1
Was reading around different articles and discussions and found something interesting. There was discussion about head replacing and a guy showed a graph picture that shows a PB FR when playing a test tape (10kHz, 400Hz and 3kHz). There was a diff about 1db between 400Hz and 3kHz. One other guy wrote that a difference between 400Hz and 3kHz can be fixed by increasing the PB EQ time constant. Graph for my deck looks about the the same (more or less) and a diff is about 1.5db.

1659895532130.png

Now I'm wondering if I could do something like that and make PB FR flat or if there is some other way to make that correction?

PS I know that frequencies should be in order 400Hz-3kHz-10kHz but a test tape is recorded as showed on that picture.

Now Vince is gonna say "I told you..." :)
 
Last edited:

vince666

Chief Journeyman
Joined
May 15, 2013
Messages
972
Location
deep south of Italy
Tagline
I will not be missed! :p
#5
more seriously... (but not too much)

to handle PB response there are a couple places/parts to think about, there at the PB EQ amp.

- input loading resistors

- input "loading" capacitors (not sure if it's the right way to call them, though)

- PB EQ time constants

more details later, when i will be writing from my computer (i am on the smartphone at now)
 
Last edited:

vince666

Chief Journeyman
Joined
May 15, 2013
Messages
972
Location
deep south of Italy
Tagline
I will not be missed! :p
#6
well... let's go a bit more in details in regards of the above matters... and will do examples referring to the Technics RS-B965 which is AlexSE's deck involved in this thread.

of course, being myself someone who never really studied electronics, my "knowledge" comes from what I could learn from other expert people (i.e. Alex/A.N.T. who i will never end to thank!) , by looking at schematics and by experimenting here at home on my decks while measuring the results... so, please, forgive me if i might say something the unproper way.

1) input loading resistors

They are those resistors which are placed just after the PB head and they are put in parallel, between the "+" of the PB head and the ground.
Input load resistance does influence the frequency response in the way that raising the resistance value raises the FR curve on the treble side but its influence on the FR is quite wide.
I usually see, on schematics, that heads with the lowest losses on treble (i.e. amorphous and ferrite) usually show lower input loading resistors at the PB amp while the heads with higher losses on treble (permalloy and even more sendust ones) have higher value resistors.
Values of such resistors usually range in the tens of thousands of ohms and can be as high as a couple hundreds of thousands of ohms or so.

2) input loading capacitors

Similarly as above, there are also small value capacitors (in the range of the hundreds of picofarads) put in parallel between the "+" of the PB head and the ground, usually just at the same place in the schematic as the above resistor.
Also here, raising the capacitor value will raise the response on treble but, this time, the range where the capacitance has its most evident effect is less wide and located just on the treble zone... also raising capacitor value will reduce the treble extension of the FR... so, i.e., if i raise the capacitor i might see a boost around 10Khz but then the response will start to fall a bit before at the extreme treble.
And, since the effect of the capacitor is less wide than what happens with the resistor, it might alter the FR in order to make it not regular (but it's also useful to fix a FR which was not regular, though).

3) PB EQ time constants

First off, the playback head itself acts like an inductor, then its own natural frequency response grows with frequency, then there are needed low pass filters to get, as a result, a flat playback frequency response.

So, we are talking about low pass filters of a given time constant which is related to the roll-off frequency.
We know the two standard PB EQ time constants are 70 and 120 microseconds (well, they are 3 because there is also the 3180us but we never touch it as it's usually OK as is on all decks).
120 is for type1 playback and 70 for type2 and type4 (unless we deliberately choose to record and play such tapes at 120).

And, to be more precise, the "120" setting is really a 120+3180 setting while the "70" setting is really a 70+3180 setting but, here, I will only focus on the 120 and 70 time constant while deliberately ignoring the 3180 one.

If you go and check the schematics on a few decks and calculate the actual time constants, you'll see they are almost never exactly 120 and 70 and, sometimes, they might also be quite far from the standard values.
Point is that the exact standard values work fine only on heads which show a FR close to the theorethycal optimum one, then the heads which are suppoused to work fine just at 120 and 70 are the ferrite and the amorphous ones.
Keeping (freely) quoting what Alex/A.N.T. shared over the years, if a ferrite or amorphous head is OK at 120 (focusing on the time constant for type1 mode, which i'll call "main time constant" , the other one will need to follow it proportionally) then a permalloy head will need a value between 130 and 140 , and a sendust head will need a value between 140 and 170.
Of course, to find out which is the best value, it's needed to test the PB FR with a proper calibration tape.

I usually see they set the time constants with resistors and capacitors... don't know if there are also other ways to do that but, here, i will refer to R and C because it's the only method I do know.
To calculate the actual time constant in microseconds, I do multiply the resistance in Kohm with the capacitance in nanofarads.


But let's go on directly with a practical example regarding the Technics RS-B965 cassette deck.

Technics RS-B965 PB EQ amp.jpg

as you might see, this schematic shows a few useful hints... this deck does exist in more versions and, in particular, the PP version has amorphous heads while the other versions have permalloy heads.
Here in my description, I will refer only to the left channel then the same would be true on the right channel through the corresponding parts.

On this deck, we can see they did choose the input loading resistor R1 as 68Kohm for any versions of the deck.
And they did choose to add also the input "loading" capacitor C1 as 150pF on all the version except the PP version with amorphous heads.

but let's go at the PB EQ time constants...

On this deck, the PB EQ time constants are handled by R15 , R19 and C5.

The "120 microseconds" time constant, for type1 playback mode on all versions which aren't the PP one, is calculated this way:

( R15 + R19 ) * C5 = ( 15 K + 10 K ) * 5.6 nF = 25 K * 5.6 nF = 140 microseconds

while, for the PP version with amorphous heads, it becomes:

( R15 + R19 ) * C5 = ( 12 K + 10 K ) * 5.6 nF = 22 K * 5.6 nF = 123.2 microseconds

and the "70 microseconds" time constant, for type2 and type4 playback mode on all versions which aren't the PP one, is calculated this way:

R15 * C5 = 15 K * 5.6 nF = 84 microseconds

while, for the PP version with amorphous heads, it becomes:

R15 * C5 = 12 K * 5.6 nF = 67.2 microseconds


please, draw your own conclusions.


I hope it helps,

Vince.
 
Last edited:

vince666

Chief Journeyman
Joined
May 15, 2013
Messages
972
Location
deep south of Italy
Tagline
I will not be missed! :p
#8

you are welcome, my friend. :)

if needed, I've learnt how to calculate PB EQ time constants on all other Technics RS-B and RS-BX decks with that AN7351K chip inside and also on older ones from the RS-M series... so, please, feel free to ask so that i will add more examples based on the schematics of such other models...

Moreover, the RS-M models (at least on those few ones i checked the schematic) do show a mistake in the way they handle the PB EQ time constants and which Alex/A.N.T. did put in evidence while explaining me how to calculate the time constants on the RS-M260... but i see the same mistake is also on the other few RS-M decks i did check.

Also, I've learnt how to do the same on the AIWA AD-F880 (and other AIWA ones with similar kind of schematic at the PB EQ amp) just while I had to modify one of my own AD-F880 by fitting amorphous heads in place of the original permalloy ones.

Cheers,

Vince.
 

Makymak

Journeyman
Joined
Dec 6, 2021
Messages
193
Location
Where the sun meets the rocks and the sea
#9
I would definitely like to add an amorphous head on one of my 965 even by chance I could get any. But with the rarity these heads come and today prices, I think this will be just high expectations!!!

An off-topic question: having both RS-B965M and RS-B965MF do you think the M (just oem permalloy heads) stands quite well against your other decks? And what about recording? I only have Technics decks (quite a few and all moded following Alex's advices!) so I can't make any comparisons.
 

vince666

Chief Journeyman
Joined
May 15, 2013
Messages
972
Location
deep south of Italy
Tagline
I will not be missed! :p
#10
I would definitely like to add an amorphous head on one of my 965 even by chance I could get any. But with the rarity these heads come and today prices, I think this will be just high expectations!!!

An off-topic question: having both RS-B965M and RS-B965MF do you think the M (just oem permalloy heads) stands quite well against your other decks? And what about recording? I only have Technics decks (quite a few and all moded following Alex's advices!) so I can't make any comparisons.
yes, the RS-B965M with its own original permalloy heads is just a fantastic deck!
On playback quality, we know it's just at the TOP and the other "MF" version with Alps ferrite heads is just as good but with some lower hiss due to the "quieter" ferrite heads compared to the permalloy ones.
Recording-wise, the RS-B965M makes a terrific job but the "MF" version makes an incredible work as it can push the tapes even more without going into distortion.

Compared to all my other decks, both of my RS-B965 modified by Alex are, of course, my own benchmark ones and the best ones I have.

Anyways, the kind of heads do influence a lot the actual performance of a deck... on playback but even more on recording, by giving you the ability to push any tapes to the limit... i.e. amorphous heads, properly exploited by the electronics, will be able to push just ANY tapes to their own limit...

As an example, my Technics RS-M260 has some nice Canon amorphous heads (originally it had some horrible sounding Canon sendust heads!)... and with the amorphous heads I could measure a MOL at 315Hz of +12.7dB VU on a Sony Metal ES from 1985.... and a higher than expected MOL on quite any other tapes, like i.e. about +9dB VU on TDK D from 1982-83... not bad at all!
But, to be sincere, also the sendust heads are able to push tapes almost as much but many (most?) of them do sound horrible on playback, to my ears... and, btw, even if the distortion performance (then the MOL@315Hz i could get) on the M260 was almost on par with both the original sendust heads and these amorphous heads (with some advantage on the amorphous heads), the original sendust heads, despite i had also lapped them, did show a very limited frequency bandwidth (not much extended on the treble side!) , also the official specs in the service manual aren't that great, in fact... while, with the amorphous heads on the same deck i did automatically gain about 6-7Khz more frequency bandwidth on the treble side on just ANY tapes... i.e. , if a Sony HF could reach only 16Khz on those original (horrible) sendust heads, the same tape can reach about 22-23Khz with the amorphous heads! And this deck doesn't even have any kind of HX-Pro!
Anyways, i suspect the not great frequency bandwidth at the treble side on the RS-M260 with original sendust heads is mostly due to the playback head than to the recording head, in fact, the tapes recorded in this deck with original heads did sound better when played on any other decks than on the same RS-M260 which originally recorded them!
(PS: and if you check the schematic of the RS-M260, they set the PB EQ time constant on the sendust heads at a value which would be OK on amorphous or ferrite heads, then i had also raised that value to fix the PB response with the original sendust heads but the sound, even if the FR was correct, was maybe even uglier than with the original PB response which shows a falling behaviour)

But, of course, even if modifying the PB side to accomodate a different kind of head is quite simple, to fix the recording side is quite a bit complex, because you should at least check/modify the bias , the REC gain and the the REC EQ filters... and, most (if not all) of the times, you just need to modify the circuit because while changing the kind of heads the tuneable trimmers don't have enough range to accomodate the new heads properly. So, changing the heads with some made of different material isn't a straightforward work (as Alex did warn me in the past, while being just right on it!) but i add that the most difficult part is to get the REC side right as the PB side is a lot easier to handle.
 
Last edited:

vince666

Chief Journeyman
Joined
May 15, 2013
Messages
972
Location
deep south of Italy
Tagline
I will not be missed! :p
#11
about the RS-B965M with original permalloy heads...

did you add the new extra mod Alex did share in the most recent years which involves the HX-PRO circuit, in order to make it work as much as needed?
 

Makymak

Journeyman
Joined
Dec 6, 2021
Messages
193
Location
Where the sun meets the rocks and the sea
#12
No, I didn't made the resistors mod on the HX pro. I need to find some time to make it on all my decks.

I know how challenging is to tune a different type of head on the recording side. That's why I'd only consider to replace the 965's head since Technics has both versions and the parts needed are well documented in the SM.
 

vince666

Chief Journeyman
Joined
May 15, 2013
Messages
972
Location
deep south of Italy
Tagline
I will not be missed! :p
#13
No, I didn't made the resistors mod on the HX pro. I need to find some time to make it on all my decks.
ah nice! so you just know all the details about what to do? (but it's a really simple and quick mod, though)

I've made it on my RS-B965M (Alex told it's not needed on the RS-B965MF, though) and I've noticed an improvement in recording quality on type1 tapes, the kind of tapes which can actually benefit from the HX-PRO... on type2 and type4 tapes the benefit of the HX-PRO isn't that noticeable as their greater coercivity just helps reducing the "self erasure" effect on treble (due to the bias while recording)... but on the type1 tapes the HX-PRO which chimes in when actually needed is surely a welcome add.

I know how challenging is to tune a different type of head on the recording side. That's why I'd only consider to replace the 965's head since Technics has both versions and the parts needed are well documented in the SM.
yes, this is a lucky case so that the SM is just full of useful hints. :)
 
Last edited:

Alex SE

Journeyman
Joined
Oct 10, 2021
Messages
246
Location
Sweden
Tagline
Amateur
#14
Update for those who are curios:
Replaced R1 (68K) with 100K and got much better (flatter) FR. Now just need to adjust bias but that's easy.

Big thanx to brother Vince for his time, patience and of course for sharing his knowledge.

BTW, anyone know if Alex A.N.T. is still selling a Test tapes? Need one combined (400Hz/3kHz/10kHz) or eventually 3 separate tapes.
 

vince666

Chief Journeyman
Joined
May 15, 2013
Messages
972
Location
deep south of Italy
Tagline
I will not be missed! :p
#17
why should you make it better than that? :p:);)

and, yes, roll off at low end is both on source and tape, then it doesn't matter... the deck is doing its work just nicely...

excellent! :)
 
Top